Upload 55 files
Browse filesThis view is limited to 50 files because it contains too many changes.
See raw diff
- fine_tuned_model_output_1/checkpoint-10000/config.json +30 -0
- fine_tuned_model_output_1/checkpoint-10000/generation_config.json +6 -0
- fine_tuned_model_output_1/checkpoint-10000/model.safetensors +3 -0
- fine_tuned_model_output_1/checkpoint-10000/optimizer.pt +3 -0
- fine_tuned_model_output_1/checkpoint-10000/rng_state.pth +3 -0
- fine_tuned_model_output_1/checkpoint-10000/scaler.pt +3 -0
- fine_tuned_model_output_1/checkpoint-10000/scheduler.pt +3 -0
- fine_tuned_model_output_1/checkpoint-10000/special_tokens_map.json +30 -0
- fine_tuned_model_output_1/checkpoint-10000/tokenizer.json +0 -0
- fine_tuned_model_output_1/checkpoint-10000/tokenizer_config.json +67 -0
- fine_tuned_model_output_1/checkpoint-10000/trainer_state.json +813 -0
- fine_tuned_model_output_1/checkpoint-10000/training_args.bin +3 -0
- fine_tuned_model_output_1/checkpoint-20000/config.json +30 -0
- fine_tuned_model_output_1/checkpoint-20000/generation_config.json +6 -0
- fine_tuned_model_output_1/checkpoint-20000/model.safetensors +3 -0
- fine_tuned_model_output_1/checkpoint-20000/optimizer.pt +3 -0
- fine_tuned_model_output_1/checkpoint-20000/rng_state.pth +3 -0
- fine_tuned_model_output_1/checkpoint-20000/scaler.pt +3 -0
- fine_tuned_model_output_1/checkpoint-20000/scheduler.pt +3 -0
- fine_tuned_model_output_1/checkpoint-20000/special_tokens_map.json +30 -0
- fine_tuned_model_output_1/checkpoint-20000/tokenizer.json +0 -0
- fine_tuned_model_output_1/checkpoint-20000/tokenizer_config.json +67 -0
- fine_tuned_model_output_1/checkpoint-20000/trainer_state.json +1593 -0
- fine_tuned_model_output_1/checkpoint-20000/training_args.bin +3 -0
- fine_tuned_model_output_1/checkpoint-30000/config.json +30 -0
- fine_tuned_model_output_1/checkpoint-30000/generation_config.json +6 -0
- fine_tuned_model_output_1/checkpoint-30000/model.safetensors +3 -0
- fine_tuned_model_output_1/checkpoint-30000/optimizer.pt +3 -0
- fine_tuned_model_output_1/checkpoint-30000/rng_state.pth +3 -0
- fine_tuned_model_output_1/checkpoint-30000/scaler.pt +3 -0
- fine_tuned_model_output_1/checkpoint-30000/scheduler.pt +3 -0
- fine_tuned_model_output_1/checkpoint-30000/special_tokens_map.json +30 -0
- fine_tuned_model_output_1/checkpoint-30000/tokenizer.json +0 -0
- fine_tuned_model_output_1/checkpoint-30000/tokenizer_config.json +67 -0
- fine_tuned_model_output_1/checkpoint-30000/trainer_state.json +2373 -0
- fine_tuned_model_output_1/checkpoint-30000/training_args.bin +3 -0
- fine_tuned_model_output_1/checkpoint-31649/config.json +30 -0
- fine_tuned_model_output_1/checkpoint-31649/generation_config.json +6 -0
- fine_tuned_model_output_1/checkpoint-31649/model.safetensors +3 -0
- fine_tuned_model_output_1/checkpoint-31649/optimizer.pt +3 -0
- fine_tuned_model_output_1/checkpoint-31649/rng_state.pth +3 -0
- fine_tuned_model_output_1/checkpoint-31649/scaler.pt +3 -0
- fine_tuned_model_output_1/checkpoint-31649/scheduler.pt +3 -0
- fine_tuned_model_output_1/checkpoint-31649/special_tokens_map.json +30 -0
- fine_tuned_model_output_1/checkpoint-31649/tokenizer.json +0 -0
- fine_tuned_model_output_1/checkpoint-31649/tokenizer_config.json +67 -0
- fine_tuned_model_output_1/checkpoint-31649/trainer_state.json +2493 -0
- fine_tuned_model_output_1/checkpoint-31649/training_args.bin +3 -0
- fine_tuned_model_output_1/config.json +30 -0
- fine_tuned_model_output_1/generation_config.json +6 -0
fine_tuned_model_output_1/checkpoint-10000/config.json
ADDED
|
@@ -0,0 +1,30 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"_name_or_path": "./linked_model_with_custom_tokenizer",
|
| 3 |
+
"architectures": [
|
| 4 |
+
"Qwen2ForCausalLM"
|
| 5 |
+
],
|
| 6 |
+
"attention_dropout": 0.0,
|
| 7 |
+
"bos_token_id": 151643,
|
| 8 |
+
"eos_token_id": 151643,
|
| 9 |
+
"hidden_act": "silu",
|
| 10 |
+
"hidden_size": 896,
|
| 11 |
+
"initializer_range": 0.02,
|
| 12 |
+
"intermediate_size": 4864,
|
| 13 |
+
"max_position_embeddings": 32768,
|
| 14 |
+
"max_window_layers": 24,
|
| 15 |
+
"model_type": "qwen2",
|
| 16 |
+
"num_attention_heads": 14,
|
| 17 |
+
"num_hidden_layers": 4,
|
| 18 |
+
"num_key_value_heads": 2,
|
| 19 |
+
"rms_norm_eps": 1e-06,
|
| 20 |
+
"rope_scaling": null,
|
| 21 |
+
"rope_theta": 1000000.0,
|
| 22 |
+
"sliding_window": 32768,
|
| 23 |
+
"tie_word_embeddings": true,
|
| 24 |
+
"torch_dtype": "float32",
|
| 25 |
+
"transformers_version": "4.49.0",
|
| 26 |
+
"use_cache": true,
|
| 27 |
+
"use_mrope": false,
|
| 28 |
+
"use_sliding_window": false,
|
| 29 |
+
"vocab_size": 32002
|
| 30 |
+
}
|
fine_tuned_model_output_1/checkpoint-10000/generation_config.json
ADDED
|
@@ -0,0 +1,6 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"_from_model_config": true,
|
| 3 |
+
"bos_token_id": 151643,
|
| 4 |
+
"eos_token_id": 151643,
|
| 5 |
+
"transformers_version": "4.49.0"
|
| 6 |
+
}
|
fine_tuned_model_output_1/checkpoint-10000/model.safetensors
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:fddd77e38534fababa57514b96fae807a7d396eb2d0d03ea6ab931768368ce98
|
| 3 |
+
size 353302384
|
fine_tuned_model_output_1/checkpoint-10000/optimizer.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:bd651167b3298d0d1e3a071fb172f27e967b7a53310a4e161941cd10496539a1
|
| 3 |
+
size 706636090
|
fine_tuned_model_output_1/checkpoint-10000/rng_state.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:c295ee6b6de2d3abd8a1fc93136e75c566269d23f9d0fe83fe1ccc60ad2f6d63
|
| 3 |
+
size 14244
|
fine_tuned_model_output_1/checkpoint-10000/scaler.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:bd9c9ba323545cf3c37ea63884860765d7331d69f7174459084dcd7856919ade
|
| 3 |
+
size 988
|
fine_tuned_model_output_1/checkpoint-10000/scheduler.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:0352d13234b6a75552ff46f42418bbf2f25a7ea01048637f44712f33eef0560e
|
| 3 |
+
size 1064
|
fine_tuned_model_output_1/checkpoint-10000/special_tokens_map.json
ADDED
|
@@ -0,0 +1,30 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"bos_token": {
|
| 3 |
+
"content": "<s>",
|
| 4 |
+
"lstrip": false,
|
| 5 |
+
"normalized": false,
|
| 6 |
+
"rstrip": false,
|
| 7 |
+
"single_word": false
|
| 8 |
+
},
|
| 9 |
+
"eos_token": {
|
| 10 |
+
"content": "</s>",
|
| 11 |
+
"lstrip": false,
|
| 12 |
+
"normalized": false,
|
| 13 |
+
"rstrip": false,
|
| 14 |
+
"single_word": false
|
| 15 |
+
},
|
| 16 |
+
"pad_token": {
|
| 17 |
+
"content": "</s>",
|
| 18 |
+
"lstrip": false,
|
| 19 |
+
"normalized": false,
|
| 20 |
+
"rstrip": false,
|
| 21 |
+
"single_word": false
|
| 22 |
+
},
|
| 23 |
+
"unk_token": {
|
| 24 |
+
"content": "<unk>",
|
| 25 |
+
"lstrip": false,
|
| 26 |
+
"normalized": false,
|
| 27 |
+
"rstrip": false,
|
| 28 |
+
"single_word": false
|
| 29 |
+
}
|
| 30 |
+
}
|
fine_tuned_model_output_1/checkpoint-10000/tokenizer.json
ADDED
|
The diff for this file is too large to render.
See raw diff
|
|
|
fine_tuned_model_output_1/checkpoint-10000/tokenizer_config.json
ADDED
|
@@ -0,0 +1,67 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"added_tokens_decoder": {
|
| 3 |
+
"0": {
|
| 4 |
+
"content": "<unk>",
|
| 5 |
+
"lstrip": false,
|
| 6 |
+
"normalized": false,
|
| 7 |
+
"rstrip": false,
|
| 8 |
+
"single_word": false,
|
| 9 |
+
"special": true
|
| 10 |
+
},
|
| 11 |
+
"1": {
|
| 12 |
+
"content": "<s>",
|
| 13 |
+
"lstrip": false,
|
| 14 |
+
"normalized": false,
|
| 15 |
+
"rstrip": false,
|
| 16 |
+
"single_word": false,
|
| 17 |
+
"special": true
|
| 18 |
+
},
|
| 19 |
+
"2": {
|
| 20 |
+
"content": "</s>",
|
| 21 |
+
"lstrip": false,
|
| 22 |
+
"normalized": false,
|
| 23 |
+
"rstrip": false,
|
| 24 |
+
"single_word": false,
|
| 25 |
+
"special": true
|
| 26 |
+
},
|
| 27 |
+
"3": {
|
| 28 |
+
"content": "<pad>",
|
| 29 |
+
"lstrip": false,
|
| 30 |
+
"normalized": false,
|
| 31 |
+
"rstrip": false,
|
| 32 |
+
"single_word": false,
|
| 33 |
+
"special": true
|
| 34 |
+
},
|
| 35 |
+
"32000": {
|
| 36 |
+
"content": "Sanskrit:",
|
| 37 |
+
"lstrip": false,
|
| 38 |
+
"normalized": true,
|
| 39 |
+
"rstrip": false,
|
| 40 |
+
"single_word": false,
|
| 41 |
+
"special": false
|
| 42 |
+
},
|
| 43 |
+
"32001": {
|
| 44 |
+
"content": "English:",
|
| 45 |
+
"lstrip": false,
|
| 46 |
+
"normalized": true,
|
| 47 |
+
"rstrip": false,
|
| 48 |
+
"single_word": false,
|
| 49 |
+
"special": false
|
| 50 |
+
}
|
| 51 |
+
},
|
| 52 |
+
"bos_token": "<s>",
|
| 53 |
+
"clean_up_tokenization_spaces": false,
|
| 54 |
+
"eos_token": "</s>",
|
| 55 |
+
"extra_special_tokens": {},
|
| 56 |
+
"max_length": 512,
|
| 57 |
+
"model_max_length": 1000000000000000019884624838656,
|
| 58 |
+
"pad_to_multiple_of": null,
|
| 59 |
+
"pad_token": "</s>",
|
| 60 |
+
"pad_token_type_id": 0,
|
| 61 |
+
"padding_side": "right",
|
| 62 |
+
"stride": 0,
|
| 63 |
+
"tokenizer_class": "PreTrainedTokenizer",
|
| 64 |
+
"truncation_side": "right",
|
| 65 |
+
"truncation_strategy": "longest_first",
|
| 66 |
+
"unk_token": "<unk>"
|
| 67 |
+
}
|
fine_tuned_model_output_1/checkpoint-10000/trainer_state.json
ADDED
|
@@ -0,0 +1,813 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": 4.511664390563965,
|
| 3 |
+
"best_model_checkpoint": "./fine_tuned_model_output_1\\checkpoint-10000",
|
| 4 |
+
"epoch": 0.3159597593650473,
|
| 5 |
+
"eval_steps": 1000,
|
| 6 |
+
"global_step": 10000,
|
| 7 |
+
"is_hyper_param_search": false,
|
| 8 |
+
"is_local_process_zero": true,
|
| 9 |
+
"is_world_process_zero": true,
|
| 10 |
+
"log_history": [
|
| 11 |
+
{
|
| 12 |
+
"epoch": 0.0031595975936504726,
|
| 13 |
+
"grad_norm": 5.00691556930542,
|
| 14 |
+
"learning_rate": 4.984517678283674e-05,
|
| 15 |
+
"loss": 5.1637,
|
| 16 |
+
"step": 100
|
| 17 |
+
},
|
| 18 |
+
{
|
| 19 |
+
"epoch": 0.006319195187300945,
|
| 20 |
+
"grad_norm": 4.790140628814697,
|
| 21 |
+
"learning_rate": 4.968719390818036e-05,
|
| 22 |
+
"loss": 4.9551,
|
| 23 |
+
"step": 200
|
| 24 |
+
},
|
| 25 |
+
{
|
| 26 |
+
"epoch": 0.009478792780951418,
|
| 27 |
+
"grad_norm": 4.75383186340332,
|
| 28 |
+
"learning_rate": 4.952921103352397e-05,
|
| 29 |
+
"loss": 4.8786,
|
| 30 |
+
"step": 300
|
| 31 |
+
},
|
| 32 |
+
{
|
| 33 |
+
"epoch": 0.01263839037460189,
|
| 34 |
+
"grad_norm": 5.218325138092041,
|
| 35 |
+
"learning_rate": 4.9371228158867584e-05,
|
| 36 |
+
"loss": 4.8117,
|
| 37 |
+
"step": 400
|
| 38 |
+
},
|
| 39 |
+
{
|
| 40 |
+
"epoch": 0.015797987968252363,
|
| 41 |
+
"grad_norm": 5.101843357086182,
|
| 42 |
+
"learning_rate": 4.9213245284211194e-05,
|
| 43 |
+
"loss": 4.8153,
|
| 44 |
+
"step": 500
|
| 45 |
+
},
|
| 46 |
+
{
|
| 47 |
+
"epoch": 0.018957585561902836,
|
| 48 |
+
"grad_norm": 4.165743350982666,
|
| 49 |
+
"learning_rate": 4.905526240955481e-05,
|
| 50 |
+
"loss": 4.7331,
|
| 51 |
+
"step": 600
|
| 52 |
+
},
|
| 53 |
+
{
|
| 54 |
+
"epoch": 0.02211718315555331,
|
| 55 |
+
"grad_norm": 4.386114597320557,
|
| 56 |
+
"learning_rate": 4.889727953489842e-05,
|
| 57 |
+
"loss": 4.7982,
|
| 58 |
+
"step": 700
|
| 59 |
+
},
|
| 60 |
+
{
|
| 61 |
+
"epoch": 0.02527678074920378,
|
| 62 |
+
"grad_norm": 4.677374362945557,
|
| 63 |
+
"learning_rate": 4.8739296660242035e-05,
|
| 64 |
+
"loss": 4.751,
|
| 65 |
+
"step": 800
|
| 66 |
+
},
|
| 67 |
+
{
|
| 68 |
+
"epoch": 0.028436378342854254,
|
| 69 |
+
"grad_norm": 4.1365814208984375,
|
| 70 |
+
"learning_rate": 4.8581313785585644e-05,
|
| 71 |
+
"loss": 4.7631,
|
| 72 |
+
"step": 900
|
| 73 |
+
},
|
| 74 |
+
{
|
| 75 |
+
"epoch": 0.031595975936504726,
|
| 76 |
+
"grad_norm": 4.651081562042236,
|
| 77 |
+
"learning_rate": 4.842333091092926e-05,
|
| 78 |
+
"loss": 4.7504,
|
| 79 |
+
"step": 1000
|
| 80 |
+
},
|
| 81 |
+
{
|
| 82 |
+
"epoch": 0.031595975936504726,
|
| 83 |
+
"eval_loss": 4.873119354248047,
|
| 84 |
+
"eval_runtime": 665.3679,
|
| 85 |
+
"eval_samples_per_second": 132.131,
|
| 86 |
+
"eval_steps_per_second": 16.517,
|
| 87 |
+
"step": 1000
|
| 88 |
+
},
|
| 89 |
+
{
|
| 90 |
+
"epoch": 0.0347555735301552,
|
| 91 |
+
"grad_norm": 5.823278903961182,
|
| 92 |
+
"learning_rate": 4.826534803627287e-05,
|
| 93 |
+
"loss": 4.6999,
|
| 94 |
+
"step": 1100
|
| 95 |
+
},
|
| 96 |
+
{
|
| 97 |
+
"epoch": 0.03791517112380567,
|
| 98 |
+
"grad_norm": 4.6261773109436035,
|
| 99 |
+
"learning_rate": 4.8107365161616486e-05,
|
| 100 |
+
"loss": 4.6704,
|
| 101 |
+
"step": 1200
|
| 102 |
+
},
|
| 103 |
+
{
|
| 104 |
+
"epoch": 0.04107476871745615,
|
| 105 |
+
"grad_norm": 4.6412200927734375,
|
| 106 |
+
"learning_rate": 4.79493822869601e-05,
|
| 107 |
+
"loss": 4.691,
|
| 108 |
+
"step": 1300
|
| 109 |
+
},
|
| 110 |
+
{
|
| 111 |
+
"epoch": 0.04423436631110662,
|
| 112 |
+
"grad_norm": 4.368412494659424,
|
| 113 |
+
"learning_rate": 4.779139941230371e-05,
|
| 114 |
+
"loss": 4.7048,
|
| 115 |
+
"step": 1400
|
| 116 |
+
},
|
| 117 |
+
{
|
| 118 |
+
"epoch": 0.04739396390475709,
|
| 119 |
+
"grad_norm": 3.9410157203674316,
|
| 120 |
+
"learning_rate": 4.763499636639389e-05,
|
| 121 |
+
"loss": 4.6213,
|
| 122 |
+
"step": 1500
|
| 123 |
+
},
|
| 124 |
+
{
|
| 125 |
+
"epoch": 0.05055356149840756,
|
| 126 |
+
"grad_norm": 5.697903633117676,
|
| 127 |
+
"learning_rate": 4.74770134917375e-05,
|
| 128 |
+
"loss": 4.6791,
|
| 129 |
+
"step": 1600
|
| 130 |
+
},
|
| 131 |
+
{
|
| 132 |
+
"epoch": 0.05371315909205804,
|
| 133 |
+
"grad_norm": 4.182404041290283,
|
| 134 |
+
"learning_rate": 4.7319030617081114e-05,
|
| 135 |
+
"loss": 4.6493,
|
| 136 |
+
"step": 1700
|
| 137 |
+
},
|
| 138 |
+
{
|
| 139 |
+
"epoch": 0.05687275668570851,
|
| 140 |
+
"grad_norm": 3.8856332302093506,
|
| 141 |
+
"learning_rate": 4.7161047742424724e-05,
|
| 142 |
+
"loss": 4.6081,
|
| 143 |
+
"step": 1800
|
| 144 |
+
},
|
| 145 |
+
{
|
| 146 |
+
"epoch": 0.06003235427935898,
|
| 147 |
+
"grad_norm": 3.778503656387329,
|
| 148 |
+
"learning_rate": 4.700306486776834e-05,
|
| 149 |
+
"loss": 4.6437,
|
| 150 |
+
"step": 1900
|
| 151 |
+
},
|
| 152 |
+
{
|
| 153 |
+
"epoch": 0.06319195187300945,
|
| 154 |
+
"grad_norm": 5.072572231292725,
|
| 155 |
+
"learning_rate": 4.684508199311195e-05,
|
| 156 |
+
"loss": 4.5587,
|
| 157 |
+
"step": 2000
|
| 158 |
+
},
|
| 159 |
+
{
|
| 160 |
+
"epoch": 0.06319195187300945,
|
| 161 |
+
"eval_loss": 4.763667106628418,
|
| 162 |
+
"eval_runtime": 665.5888,
|
| 163 |
+
"eval_samples_per_second": 132.088,
|
| 164 |
+
"eval_steps_per_second": 16.512,
|
| 165 |
+
"step": 2000
|
| 166 |
+
},
|
| 167 |
+
{
|
| 168 |
+
"epoch": 0.06635154946665993,
|
| 169 |
+
"grad_norm": 5.395784854888916,
|
| 170 |
+
"learning_rate": 4.6687099118455565e-05,
|
| 171 |
+
"loss": 4.5621,
|
| 172 |
+
"step": 2100
|
| 173 |
+
},
|
| 174 |
+
{
|
| 175 |
+
"epoch": 0.0695111470603104,
|
| 176 |
+
"grad_norm": 4.856666088104248,
|
| 177 |
+
"learning_rate": 4.6529116243799174e-05,
|
| 178 |
+
"loss": 4.5721,
|
| 179 |
+
"step": 2200
|
| 180 |
+
},
|
| 181 |
+
{
|
| 182 |
+
"epoch": 0.07267074465396087,
|
| 183 |
+
"grad_norm": 4.935929298400879,
|
| 184 |
+
"learning_rate": 4.637113336914279e-05,
|
| 185 |
+
"loss": 4.6004,
|
| 186 |
+
"step": 2300
|
| 187 |
+
},
|
| 188 |
+
{
|
| 189 |
+
"epoch": 0.07583034224761134,
|
| 190 |
+
"grad_norm": 5.257358551025391,
|
| 191 |
+
"learning_rate": 4.62131504944864e-05,
|
| 192 |
+
"loss": 4.6032,
|
| 193 |
+
"step": 2400
|
| 194 |
+
},
|
| 195 |
+
{
|
| 196 |
+
"epoch": 0.07898993984126182,
|
| 197 |
+
"grad_norm": 4.67532205581665,
|
| 198 |
+
"learning_rate": 4.6055167619830016e-05,
|
| 199 |
+
"loss": 4.5734,
|
| 200 |
+
"step": 2500
|
| 201 |
+
},
|
| 202 |
+
{
|
| 203 |
+
"epoch": 0.0821495374349123,
|
| 204 |
+
"grad_norm": 4.023921966552734,
|
| 205 |
+
"learning_rate": 4.5897184745173625e-05,
|
| 206 |
+
"loss": 4.5888,
|
| 207 |
+
"step": 2600
|
| 208 |
+
},
|
| 209 |
+
{
|
| 210 |
+
"epoch": 0.08530913502856276,
|
| 211 |
+
"grad_norm": 3.9619414806365967,
|
| 212 |
+
"learning_rate": 4.573920187051724e-05,
|
| 213 |
+
"loss": 4.5703,
|
| 214 |
+
"step": 2700
|
| 215 |
+
},
|
| 216 |
+
{
|
| 217 |
+
"epoch": 0.08846873262221323,
|
| 218 |
+
"grad_norm": 3.7340047359466553,
|
| 219 |
+
"learning_rate": 4.558121899586085e-05,
|
| 220 |
+
"loss": 4.5437,
|
| 221 |
+
"step": 2800
|
| 222 |
+
},
|
| 223 |
+
{
|
| 224 |
+
"epoch": 0.09162833021586371,
|
| 225 |
+
"grad_norm": 3.5358519554138184,
|
| 226 |
+
"learning_rate": 4.5423236121204466e-05,
|
| 227 |
+
"loss": 4.5433,
|
| 228 |
+
"step": 2900
|
| 229 |
+
},
|
| 230 |
+
{
|
| 231 |
+
"epoch": 0.09478792780951419,
|
| 232 |
+
"grad_norm": 5.275386810302734,
|
| 233 |
+
"learning_rate": 4.5265253246548075e-05,
|
| 234 |
+
"loss": 4.5315,
|
| 235 |
+
"step": 3000
|
| 236 |
+
},
|
| 237 |
+
{
|
| 238 |
+
"epoch": 0.09478792780951419,
|
| 239 |
+
"eval_loss": 4.6958231925964355,
|
| 240 |
+
"eval_runtime": 660.1374,
|
| 241 |
+
"eval_samples_per_second": 133.178,
|
| 242 |
+
"eval_steps_per_second": 16.648,
|
| 243 |
+
"step": 3000
|
| 244 |
+
},
|
| 245 |
+
{
|
| 246 |
+
"epoch": 0.09794752540316465,
|
| 247 |
+
"grad_norm": 4.195945739746094,
|
| 248 |
+
"learning_rate": 4.510727037189169e-05,
|
| 249 |
+
"loss": 4.5292,
|
| 250 |
+
"step": 3100
|
| 251 |
+
},
|
| 252 |
+
{
|
| 253 |
+
"epoch": 0.10110712299681512,
|
| 254 |
+
"grad_norm": 4.976147651672363,
|
| 255 |
+
"learning_rate": 4.49492874972353e-05,
|
| 256 |
+
"loss": 4.5718,
|
| 257 |
+
"step": 3200
|
| 258 |
+
},
|
| 259 |
+
{
|
| 260 |
+
"epoch": 0.1042667205904656,
|
| 261 |
+
"grad_norm": 3.999481678009033,
|
| 262 |
+
"learning_rate": 4.479130462257892e-05,
|
| 263 |
+
"loss": 4.5177,
|
| 264 |
+
"step": 3300
|
| 265 |
+
},
|
| 266 |
+
{
|
| 267 |
+
"epoch": 0.10742631818411608,
|
| 268 |
+
"grad_norm": 5.00895357131958,
|
| 269 |
+
"learning_rate": 4.463332174792253e-05,
|
| 270 |
+
"loss": 4.5418,
|
| 271 |
+
"step": 3400
|
| 272 |
+
},
|
| 273 |
+
{
|
| 274 |
+
"epoch": 0.11058591577776654,
|
| 275 |
+
"grad_norm": 4.499265193939209,
|
| 276 |
+
"learning_rate": 4.447533887326614e-05,
|
| 277 |
+
"loss": 4.5262,
|
| 278 |
+
"step": 3500
|
| 279 |
+
},
|
| 280 |
+
{
|
| 281 |
+
"epoch": 0.11374551337141701,
|
| 282 |
+
"grad_norm": 4.686629772186279,
|
| 283 |
+
"learning_rate": 4.431735599860976e-05,
|
| 284 |
+
"loss": 4.5276,
|
| 285 |
+
"step": 3600
|
| 286 |
+
},
|
| 287 |
+
{
|
| 288 |
+
"epoch": 0.11690511096506749,
|
| 289 |
+
"grad_norm": 5.327017784118652,
|
| 290 |
+
"learning_rate": 4.415937312395337e-05,
|
| 291 |
+
"loss": 4.4939,
|
| 292 |
+
"step": 3700
|
| 293 |
+
},
|
| 294 |
+
{
|
| 295 |
+
"epoch": 0.12006470855871797,
|
| 296 |
+
"grad_norm": 3.8358302116394043,
|
| 297 |
+
"learning_rate": 4.400139024929698e-05,
|
| 298 |
+
"loss": 4.5028,
|
| 299 |
+
"step": 3800
|
| 300 |
+
},
|
| 301 |
+
{
|
| 302 |
+
"epoch": 0.12322430615236843,
|
| 303 |
+
"grad_norm": 4.16530704498291,
|
| 304 |
+
"learning_rate": 4.384340737464059e-05,
|
| 305 |
+
"loss": 4.4964,
|
| 306 |
+
"step": 3900
|
| 307 |
+
},
|
| 308 |
+
{
|
| 309 |
+
"epoch": 0.1263839037460189,
|
| 310 |
+
"grad_norm": 4.142209053039551,
|
| 311 |
+
"learning_rate": 4.36854244999842e-05,
|
| 312 |
+
"loss": 4.4899,
|
| 313 |
+
"step": 4000
|
| 314 |
+
},
|
| 315 |
+
{
|
| 316 |
+
"epoch": 0.1263839037460189,
|
| 317 |
+
"eval_loss": 4.649657249450684,
|
| 318 |
+
"eval_runtime": 689.9636,
|
| 319 |
+
"eval_samples_per_second": 127.421,
|
| 320 |
+
"eval_steps_per_second": 15.928,
|
| 321 |
+
"step": 4000
|
| 322 |
+
},
|
| 323 |
+
{
|
| 324 |
+
"epoch": 0.12954350133966938,
|
| 325 |
+
"grad_norm": 3.615328311920166,
|
| 326 |
+
"learning_rate": 4.352744162532781e-05,
|
| 327 |
+
"loss": 4.4671,
|
| 328 |
+
"step": 4100
|
| 329 |
+
},
|
| 330 |
+
{
|
| 331 |
+
"epoch": 0.13270309893331986,
|
| 332 |
+
"grad_norm": 5.139493465423584,
|
| 333 |
+
"learning_rate": 4.336945875067143e-05,
|
| 334 |
+
"loss": 4.4954,
|
| 335 |
+
"step": 4200
|
| 336 |
+
},
|
| 337 |
+
{
|
| 338 |
+
"epoch": 0.13586269652697033,
|
| 339 |
+
"grad_norm": 4.381932258605957,
|
| 340 |
+
"learning_rate": 4.3211475876015036e-05,
|
| 341 |
+
"loss": 4.5103,
|
| 342 |
+
"step": 4300
|
| 343 |
+
},
|
| 344 |
+
{
|
| 345 |
+
"epoch": 0.1390222941206208,
|
| 346 |
+
"grad_norm": 4.257575988769531,
|
| 347 |
+
"learning_rate": 4.305349300135865e-05,
|
| 348 |
+
"loss": 4.4677,
|
| 349 |
+
"step": 4400
|
| 350 |
+
},
|
| 351 |
+
{
|
| 352 |
+
"epoch": 0.14218189171427126,
|
| 353 |
+
"grad_norm": 3.5805318355560303,
|
| 354 |
+
"learning_rate": 4.289551012670227e-05,
|
| 355 |
+
"loss": 4.4547,
|
| 356 |
+
"step": 4500
|
| 357 |
+
},
|
| 358 |
+
{
|
| 359 |
+
"epoch": 0.14534148930792173,
|
| 360 |
+
"grad_norm": 4.602927207946777,
|
| 361 |
+
"learning_rate": 4.273752725204588e-05,
|
| 362 |
+
"loss": 4.4885,
|
| 363 |
+
"step": 4600
|
| 364 |
+
},
|
| 365 |
+
{
|
| 366 |
+
"epoch": 0.1485010869015722,
|
| 367 |
+
"grad_norm": 4.054294586181641,
|
| 368 |
+
"learning_rate": 4.2579544377389494e-05,
|
| 369 |
+
"loss": 4.4791,
|
| 370 |
+
"step": 4700
|
| 371 |
+
},
|
| 372 |
+
{
|
| 373 |
+
"epoch": 0.1516606844952227,
|
| 374 |
+
"grad_norm": 3.857389450073242,
|
| 375 |
+
"learning_rate": 4.24215615027331e-05,
|
| 376 |
+
"loss": 4.4316,
|
| 377 |
+
"step": 4800
|
| 378 |
+
},
|
| 379 |
+
{
|
| 380 |
+
"epoch": 0.15482028208887316,
|
| 381 |
+
"grad_norm": 3.913726806640625,
|
| 382 |
+
"learning_rate": 4.226357862807672e-05,
|
| 383 |
+
"loss": 4.4786,
|
| 384 |
+
"step": 4900
|
| 385 |
+
},
|
| 386 |
+
{
|
| 387 |
+
"epoch": 0.15797987968252364,
|
| 388 |
+
"grad_norm": 4.159733295440674,
|
| 389 |
+
"learning_rate": 4.210559575342033e-05,
|
| 390 |
+
"loss": 4.422,
|
| 391 |
+
"step": 5000
|
| 392 |
+
},
|
| 393 |
+
{
|
| 394 |
+
"epoch": 0.15797987968252364,
|
| 395 |
+
"eval_loss": 4.6132493019104,
|
| 396 |
+
"eval_runtime": 721.543,
|
| 397 |
+
"eval_samples_per_second": 121.844,
|
| 398 |
+
"eval_steps_per_second": 15.231,
|
| 399 |
+
"step": 5000
|
| 400 |
+
},
|
| 401 |
+
{
|
| 402 |
+
"epoch": 0.16113947727617411,
|
| 403 |
+
"grad_norm": 5.403946876525879,
|
| 404 |
+
"learning_rate": 4.1947612878763944e-05,
|
| 405 |
+
"loss": 4.4857,
|
| 406 |
+
"step": 5100
|
| 407 |
+
},
|
| 408 |
+
{
|
| 409 |
+
"epoch": 0.1642990748698246,
|
| 410 |
+
"grad_norm": 11.68526554107666,
|
| 411 |
+
"learning_rate": 4.178963000410755e-05,
|
| 412 |
+
"loss": 4.4777,
|
| 413 |
+
"step": 5200
|
| 414 |
+
},
|
| 415 |
+
{
|
| 416 |
+
"epoch": 0.16745867246347504,
|
| 417 |
+
"grad_norm": 4.937869548797607,
|
| 418 |
+
"learning_rate": 4.163164712945117e-05,
|
| 419 |
+
"loss": 4.4452,
|
| 420 |
+
"step": 5300
|
| 421 |
+
},
|
| 422 |
+
{
|
| 423 |
+
"epoch": 0.17061827005712552,
|
| 424 |
+
"grad_norm": 3.7432169914245605,
|
| 425 |
+
"learning_rate": 4.147366425479478e-05,
|
| 426 |
+
"loss": 4.4717,
|
| 427 |
+
"step": 5400
|
| 428 |
+
},
|
| 429 |
+
{
|
| 430 |
+
"epoch": 0.173777867650776,
|
| 431 |
+
"grad_norm": 3.801788806915283,
|
| 432 |
+
"learning_rate": 4.1317261208884964e-05,
|
| 433 |
+
"loss": 4.4424,
|
| 434 |
+
"step": 5500
|
| 435 |
+
},
|
| 436 |
+
{
|
| 437 |
+
"epoch": 0.17693746524442647,
|
| 438 |
+
"grad_norm": 4.984846115112305,
|
| 439 |
+
"learning_rate": 4.115927833422857e-05,
|
| 440 |
+
"loss": 4.4671,
|
| 441 |
+
"step": 5600
|
| 442 |
+
},
|
| 443 |
+
{
|
| 444 |
+
"epoch": 0.18009706283807694,
|
| 445 |
+
"grad_norm": 3.8125479221343994,
|
| 446 |
+
"learning_rate": 4.100287528831875e-05,
|
| 447 |
+
"loss": 4.4079,
|
| 448 |
+
"step": 5700
|
| 449 |
+
},
|
| 450 |
+
{
|
| 451 |
+
"epoch": 0.18325666043172742,
|
| 452 |
+
"grad_norm": 5.528467655181885,
|
| 453 |
+
"learning_rate": 4.084489241366236e-05,
|
| 454 |
+
"loss": 4.4194,
|
| 455 |
+
"step": 5800
|
| 456 |
+
},
|
| 457 |
+
{
|
| 458 |
+
"epoch": 0.1864162580253779,
|
| 459 |
+
"grad_norm": 4.0301899909973145,
|
| 460 |
+
"learning_rate": 4.068690953900598e-05,
|
| 461 |
+
"loss": 4.4299,
|
| 462 |
+
"step": 5900
|
| 463 |
+
},
|
| 464 |
+
{
|
| 465 |
+
"epoch": 0.18957585561902837,
|
| 466 |
+
"grad_norm": 3.4992246627807617,
|
| 467 |
+
"learning_rate": 4.0528926664349586e-05,
|
| 468 |
+
"loss": 4.448,
|
| 469 |
+
"step": 6000
|
| 470 |
+
},
|
| 471 |
+
{
|
| 472 |
+
"epoch": 0.18957585561902837,
|
| 473 |
+
"eval_loss": 4.5857086181640625,
|
| 474 |
+
"eval_runtime": 702.1155,
|
| 475 |
+
"eval_samples_per_second": 125.216,
|
| 476 |
+
"eval_steps_per_second": 15.653,
|
| 477 |
+
"step": 6000
|
| 478 |
+
},
|
| 479 |
+
{
|
| 480 |
+
"epoch": 0.19273545321267882,
|
| 481 |
+
"grad_norm": 4.121975421905518,
|
| 482 |
+
"learning_rate": 4.03709437896932e-05,
|
| 483 |
+
"loss": 4.4076,
|
| 484 |
+
"step": 6100
|
| 485 |
+
},
|
| 486 |
+
{
|
| 487 |
+
"epoch": 0.1958950508063293,
|
| 488 |
+
"grad_norm": 3.662929058074951,
|
| 489 |
+
"learning_rate": 4.021296091503681e-05,
|
| 490 |
+
"loss": 4.3953,
|
| 491 |
+
"step": 6200
|
| 492 |
+
},
|
| 493 |
+
{
|
| 494 |
+
"epoch": 0.19905464839997977,
|
| 495 |
+
"grad_norm": 4.008944988250732,
|
| 496 |
+
"learning_rate": 4.005497804038043e-05,
|
| 497 |
+
"loss": 4.4581,
|
| 498 |
+
"step": 6300
|
| 499 |
+
},
|
| 500 |
+
{
|
| 501 |
+
"epoch": 0.20221424599363025,
|
| 502 |
+
"grad_norm": 3.7961764335632324,
|
| 503 |
+
"learning_rate": 3.9896995165724036e-05,
|
| 504 |
+
"loss": 4.397,
|
| 505 |
+
"step": 6400
|
| 506 |
+
},
|
| 507 |
+
{
|
| 508 |
+
"epoch": 0.20537384358728072,
|
| 509 |
+
"grad_norm": 4.710026741027832,
|
| 510 |
+
"learning_rate": 3.973901229106765e-05,
|
| 511 |
+
"loss": 4.4515,
|
| 512 |
+
"step": 6500
|
| 513 |
+
},
|
| 514 |
+
{
|
| 515 |
+
"epoch": 0.2085334411809312,
|
| 516 |
+
"grad_norm": 4.220251083374023,
|
| 517 |
+
"learning_rate": 3.958102941641126e-05,
|
| 518 |
+
"loss": 4.4199,
|
| 519 |
+
"step": 6600
|
| 520 |
+
},
|
| 521 |
+
{
|
| 522 |
+
"epoch": 0.21169303877458168,
|
| 523 |
+
"grad_norm": 3.7243406772613525,
|
| 524 |
+
"learning_rate": 3.942304654175488e-05,
|
| 525 |
+
"loss": 4.4097,
|
| 526 |
+
"step": 6700
|
| 527 |
+
},
|
| 528 |
+
{
|
| 529 |
+
"epoch": 0.21485263636823215,
|
| 530 |
+
"grad_norm": 4.340923309326172,
|
| 531 |
+
"learning_rate": 3.926664349584505e-05,
|
| 532 |
+
"loss": 4.4105,
|
| 533 |
+
"step": 6800
|
| 534 |
+
},
|
| 535 |
+
{
|
| 536 |
+
"epoch": 0.2180122339618826,
|
| 537 |
+
"grad_norm": 4.304684638977051,
|
| 538 |
+
"learning_rate": 3.9108660621188665e-05,
|
| 539 |
+
"loss": 4.4033,
|
| 540 |
+
"step": 6900
|
| 541 |
+
},
|
| 542 |
+
{
|
| 543 |
+
"epoch": 0.22117183155553308,
|
| 544 |
+
"grad_norm": 4.587896347045898,
|
| 545 |
+
"learning_rate": 3.8950677746532275e-05,
|
| 546 |
+
"loss": 4.4451,
|
| 547 |
+
"step": 7000
|
| 548 |
+
},
|
| 549 |
+
{
|
| 550 |
+
"epoch": 0.22117183155553308,
|
| 551 |
+
"eval_loss": 4.5615153312683105,
|
| 552 |
+
"eval_runtime": 695.974,
|
| 553 |
+
"eval_samples_per_second": 126.321,
|
| 554 |
+
"eval_steps_per_second": 15.791,
|
| 555 |
+
"step": 7000
|
| 556 |
+
},
|
| 557 |
+
{
|
| 558 |
+
"epoch": 0.22433142914918355,
|
| 559 |
+
"grad_norm": 4.658904075622559,
|
| 560 |
+
"learning_rate": 3.879269487187589e-05,
|
| 561 |
+
"loss": 4.407,
|
| 562 |
+
"step": 7100
|
| 563 |
+
},
|
| 564 |
+
{
|
| 565 |
+
"epoch": 0.22749102674283403,
|
| 566 |
+
"grad_norm": 4.226349830627441,
|
| 567 |
+
"learning_rate": 3.86347119972195e-05,
|
| 568 |
+
"loss": 4.4055,
|
| 569 |
+
"step": 7200
|
| 570 |
+
},
|
| 571 |
+
{
|
| 572 |
+
"epoch": 0.2306506243364845,
|
| 573 |
+
"grad_norm": 5.705229759216309,
|
| 574 |
+
"learning_rate": 3.8476729122563116e-05,
|
| 575 |
+
"loss": 4.3862,
|
| 576 |
+
"step": 7300
|
| 577 |
+
},
|
| 578 |
+
{
|
| 579 |
+
"epoch": 0.23381022193013498,
|
| 580 |
+
"grad_norm": 3.9644031524658203,
|
| 581 |
+
"learning_rate": 3.8318746247906725e-05,
|
| 582 |
+
"loss": 4.3647,
|
| 583 |
+
"step": 7400
|
| 584 |
+
},
|
| 585 |
+
{
|
| 586 |
+
"epoch": 0.23696981952378546,
|
| 587 |
+
"grad_norm": 4.256899356842041,
|
| 588 |
+
"learning_rate": 3.816076337325034e-05,
|
| 589 |
+
"loss": 4.4091,
|
| 590 |
+
"step": 7500
|
| 591 |
+
},
|
| 592 |
+
{
|
| 593 |
+
"epoch": 0.24012941711743593,
|
| 594 |
+
"grad_norm": 3.951305627822876,
|
| 595 |
+
"learning_rate": 3.800278049859396e-05,
|
| 596 |
+
"loss": 4.3994,
|
| 597 |
+
"step": 7600
|
| 598 |
+
},
|
| 599 |
+
{
|
| 600 |
+
"epoch": 0.2432890147110864,
|
| 601 |
+
"grad_norm": 3.6965837478637695,
|
| 602 |
+
"learning_rate": 3.7844797623937566e-05,
|
| 603 |
+
"loss": 4.4043,
|
| 604 |
+
"step": 7700
|
| 605 |
+
},
|
| 606 |
+
{
|
| 607 |
+
"epoch": 0.24644861230473686,
|
| 608 |
+
"grad_norm": 4.412446022033691,
|
| 609 |
+
"learning_rate": 3.768681474928118e-05,
|
| 610 |
+
"loss": 4.4237,
|
| 611 |
+
"step": 7800
|
| 612 |
+
},
|
| 613 |
+
{
|
| 614 |
+
"epoch": 0.24960820989838733,
|
| 615 |
+
"grad_norm": 4.190601348876953,
|
| 616 |
+
"learning_rate": 3.752883187462479e-05,
|
| 617 |
+
"loss": 4.365,
|
| 618 |
+
"step": 7900
|
| 619 |
+
},
|
| 620 |
+
{
|
| 621 |
+
"epoch": 0.2527678074920378,
|
| 622 |
+
"grad_norm": 3.6600217819213867,
|
| 623 |
+
"learning_rate": 3.737084899996841e-05,
|
| 624 |
+
"loss": 4.393,
|
| 625 |
+
"step": 8000
|
| 626 |
+
},
|
| 627 |
+
{
|
| 628 |
+
"epoch": 0.2527678074920378,
|
| 629 |
+
"eval_loss": 4.542718887329102,
|
| 630 |
+
"eval_runtime": 692.645,
|
| 631 |
+
"eval_samples_per_second": 126.928,
|
| 632 |
+
"eval_steps_per_second": 15.867,
|
| 633 |
+
"step": 8000
|
| 634 |
+
},
|
| 635 |
+
{
|
| 636 |
+
"epoch": 0.2559274050856883,
|
| 637 |
+
"grad_norm": 6.242320537567139,
|
| 638 |
+
"learning_rate": 3.721286612531202e-05,
|
| 639 |
+
"loss": 4.4185,
|
| 640 |
+
"step": 8100
|
| 641 |
+
},
|
| 642 |
+
{
|
| 643 |
+
"epoch": 0.25908700267933876,
|
| 644 |
+
"grad_norm": 4.762045383453369,
|
| 645 |
+
"learning_rate": 3.705488325065563e-05,
|
| 646 |
+
"loss": 4.3656,
|
| 647 |
+
"step": 8200
|
| 648 |
+
},
|
| 649 |
+
{
|
| 650 |
+
"epoch": 0.26224660027298924,
|
| 651 |
+
"grad_norm": 5.066886901855469,
|
| 652 |
+
"learning_rate": 3.689690037599924e-05,
|
| 653 |
+
"loss": 4.394,
|
| 654 |
+
"step": 8300
|
| 655 |
+
},
|
| 656 |
+
{
|
| 657 |
+
"epoch": 0.2654061978666397,
|
| 658 |
+
"grad_norm": 4.975124835968018,
|
| 659 |
+
"learning_rate": 3.673891750134286e-05,
|
| 660 |
+
"loss": 4.3346,
|
| 661 |
+
"step": 8400
|
| 662 |
+
},
|
| 663 |
+
{
|
| 664 |
+
"epoch": 0.2685657954602902,
|
| 665 |
+
"grad_norm": 4.436885356903076,
|
| 666 |
+
"learning_rate": 3.658093462668647e-05,
|
| 667 |
+
"loss": 4.3958,
|
| 668 |
+
"step": 8500
|
| 669 |
+
},
|
| 670 |
+
{
|
| 671 |
+
"epoch": 0.27172539305394067,
|
| 672 |
+
"grad_norm": 4.618080139160156,
|
| 673 |
+
"learning_rate": 3.6422951752030084e-05,
|
| 674 |
+
"loss": 4.3791,
|
| 675 |
+
"step": 8600
|
| 676 |
+
},
|
| 677 |
+
{
|
| 678 |
+
"epoch": 0.27488499064759114,
|
| 679 |
+
"grad_norm": 4.878961086273193,
|
| 680 |
+
"learning_rate": 3.626496887737369e-05,
|
| 681 |
+
"loss": 4.3738,
|
| 682 |
+
"step": 8700
|
| 683 |
+
},
|
| 684 |
+
{
|
| 685 |
+
"epoch": 0.2780445882412416,
|
| 686 |
+
"grad_norm": 3.7712392807006836,
|
| 687 |
+
"learning_rate": 3.610698600271731e-05,
|
| 688 |
+
"loss": 4.362,
|
| 689 |
+
"step": 8800
|
| 690 |
+
},
|
| 691 |
+
{
|
| 692 |
+
"epoch": 0.2812041858348921,
|
| 693 |
+
"grad_norm": 4.318520545959473,
|
| 694 |
+
"learning_rate": 3.594900312806092e-05,
|
| 695 |
+
"loss": 4.367,
|
| 696 |
+
"step": 8900
|
| 697 |
+
},
|
| 698 |
+
{
|
| 699 |
+
"epoch": 0.2843637834285425,
|
| 700 |
+
"grad_norm": 3.3742518424987793,
|
| 701 |
+
"learning_rate": 3.5791020253404534e-05,
|
| 702 |
+
"loss": 4.3763,
|
| 703 |
+
"step": 9000
|
| 704 |
+
},
|
| 705 |
+
{
|
| 706 |
+
"epoch": 0.2843637834285425,
|
| 707 |
+
"eval_loss": 4.525923728942871,
|
| 708 |
+
"eval_runtime": 692.5212,
|
| 709 |
+
"eval_samples_per_second": 126.951,
|
| 710 |
+
"eval_steps_per_second": 15.87,
|
| 711 |
+
"step": 9000
|
| 712 |
+
},
|
| 713 |
+
{
|
| 714 |
+
"epoch": 0.287523381022193,
|
| 715 |
+
"grad_norm": 4.027231693267822,
|
| 716 |
+
"learning_rate": 3.5633037378748143e-05,
|
| 717 |
+
"loss": 4.4086,
|
| 718 |
+
"step": 9100
|
| 719 |
+
},
|
| 720 |
+
{
|
| 721 |
+
"epoch": 0.29068297861584347,
|
| 722 |
+
"grad_norm": 4.556132793426514,
|
| 723 |
+
"learning_rate": 3.547505450409176e-05,
|
| 724 |
+
"loss": 4.3686,
|
| 725 |
+
"step": 9200
|
| 726 |
+
},
|
| 727 |
+
{
|
| 728 |
+
"epoch": 0.29384257620949394,
|
| 729 |
+
"grad_norm": 3.9203240871429443,
|
| 730 |
+
"learning_rate": 3.531707162943537e-05,
|
| 731 |
+
"loss": 4.365,
|
| 732 |
+
"step": 9300
|
| 733 |
+
},
|
| 734 |
+
{
|
| 735 |
+
"epoch": 0.2970021738031444,
|
| 736 |
+
"grad_norm": 4.094367027282715,
|
| 737 |
+
"learning_rate": 3.5159088754778985e-05,
|
| 738 |
+
"loss": 4.3734,
|
| 739 |
+
"step": 9400
|
| 740 |
+
},
|
| 741 |
+
{
|
| 742 |
+
"epoch": 0.3001617713967949,
|
| 743 |
+
"grad_norm": 5.018619060516357,
|
| 744 |
+
"learning_rate": 3.50011058801226e-05,
|
| 745 |
+
"loss": 4.4014,
|
| 746 |
+
"step": 9500
|
| 747 |
+
},
|
| 748 |
+
{
|
| 749 |
+
"epoch": 0.3033213689904454,
|
| 750 |
+
"grad_norm": 3.928473711013794,
|
| 751 |
+
"learning_rate": 3.484312300546621e-05,
|
| 752 |
+
"loss": 4.3508,
|
| 753 |
+
"step": 9600
|
| 754 |
+
},
|
| 755 |
+
{
|
| 756 |
+
"epoch": 0.30648096658409585,
|
| 757 |
+
"grad_norm": 3.889299154281616,
|
| 758 |
+
"learning_rate": 3.4685140130809826e-05,
|
| 759 |
+
"loss": 4.365,
|
| 760 |
+
"step": 9700
|
| 761 |
+
},
|
| 762 |
+
{
|
| 763 |
+
"epoch": 0.3096405641777463,
|
| 764 |
+
"grad_norm": 4.932437896728516,
|
| 765 |
+
"learning_rate": 3.4527157256153435e-05,
|
| 766 |
+
"loss": 4.3166,
|
| 767 |
+
"step": 9800
|
| 768 |
+
},
|
| 769 |
+
{
|
| 770 |
+
"epoch": 0.3128001617713968,
|
| 771 |
+
"grad_norm": 4.175967216491699,
|
| 772 |
+
"learning_rate": 3.436917438149705e-05,
|
| 773 |
+
"loss": 4.3463,
|
| 774 |
+
"step": 9900
|
| 775 |
+
},
|
| 776 |
+
{
|
| 777 |
+
"epoch": 0.3159597593650473,
|
| 778 |
+
"grad_norm": 5.54374361038208,
|
| 779 |
+
"learning_rate": 3.421119150684066e-05,
|
| 780 |
+
"loss": 4.3506,
|
| 781 |
+
"step": 10000
|
| 782 |
+
},
|
| 783 |
+
{
|
| 784 |
+
"epoch": 0.3159597593650473,
|
| 785 |
+
"eval_loss": 4.511664390563965,
|
| 786 |
+
"eval_runtime": 691.9773,
|
| 787 |
+
"eval_samples_per_second": 127.05,
|
| 788 |
+
"eval_steps_per_second": 15.882,
|
| 789 |
+
"step": 10000
|
| 790 |
+
}
|
| 791 |
+
],
|
| 792 |
+
"logging_steps": 100,
|
| 793 |
+
"max_steps": 31649,
|
| 794 |
+
"num_input_tokens_seen": 0,
|
| 795 |
+
"num_train_epochs": 1,
|
| 796 |
+
"save_steps": 10000,
|
| 797 |
+
"stateful_callbacks": {
|
| 798 |
+
"TrainerControl": {
|
| 799 |
+
"args": {
|
| 800 |
+
"should_epoch_stop": false,
|
| 801 |
+
"should_evaluate": false,
|
| 802 |
+
"should_log": false,
|
| 803 |
+
"should_save": true,
|
| 804 |
+
"should_training_stop": false
|
| 805 |
+
},
|
| 806 |
+
"attributes": {}
|
| 807 |
+
}
|
| 808 |
+
},
|
| 809 |
+
"total_flos": 4.5811531776e+16,
|
| 810 |
+
"train_batch_size": 5,
|
| 811 |
+
"trial_name": null,
|
| 812 |
+
"trial_params": null
|
| 813 |
+
}
|
fine_tuned_model_output_1/checkpoint-10000/training_args.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:95acebf5e801652183ac4ad483ff8bfe2ede739795ede7240bbee3b5a7707481
|
| 3 |
+
size 5368
|
fine_tuned_model_output_1/checkpoint-20000/config.json
ADDED
|
@@ -0,0 +1,30 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"_name_or_path": "./linked_model_with_custom_tokenizer",
|
| 3 |
+
"architectures": [
|
| 4 |
+
"Qwen2ForCausalLM"
|
| 5 |
+
],
|
| 6 |
+
"attention_dropout": 0.0,
|
| 7 |
+
"bos_token_id": 151643,
|
| 8 |
+
"eos_token_id": 151643,
|
| 9 |
+
"hidden_act": "silu",
|
| 10 |
+
"hidden_size": 896,
|
| 11 |
+
"initializer_range": 0.02,
|
| 12 |
+
"intermediate_size": 4864,
|
| 13 |
+
"max_position_embeddings": 32768,
|
| 14 |
+
"max_window_layers": 24,
|
| 15 |
+
"model_type": "qwen2",
|
| 16 |
+
"num_attention_heads": 14,
|
| 17 |
+
"num_hidden_layers": 4,
|
| 18 |
+
"num_key_value_heads": 2,
|
| 19 |
+
"rms_norm_eps": 1e-06,
|
| 20 |
+
"rope_scaling": null,
|
| 21 |
+
"rope_theta": 1000000.0,
|
| 22 |
+
"sliding_window": 32768,
|
| 23 |
+
"tie_word_embeddings": true,
|
| 24 |
+
"torch_dtype": "float32",
|
| 25 |
+
"transformers_version": "4.49.0",
|
| 26 |
+
"use_cache": true,
|
| 27 |
+
"use_mrope": false,
|
| 28 |
+
"use_sliding_window": false,
|
| 29 |
+
"vocab_size": 32002
|
| 30 |
+
}
|
fine_tuned_model_output_1/checkpoint-20000/generation_config.json
ADDED
|
@@ -0,0 +1,6 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"_from_model_config": true,
|
| 3 |
+
"bos_token_id": 151643,
|
| 4 |
+
"eos_token_id": 151643,
|
| 5 |
+
"transformers_version": "4.49.0"
|
| 6 |
+
}
|
fine_tuned_model_output_1/checkpoint-20000/model.safetensors
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:7cc0b2de1761eb02dea1fab833313a68004793803e80cc588fe68f6f0f4bbea5
|
| 3 |
+
size 353302384
|
fine_tuned_model_output_1/checkpoint-20000/optimizer.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:db0ad1f7ae47d5be9dab1284d9868e796bf44dde7248be06d4ff4c68a4c59559
|
| 3 |
+
size 706636090
|
fine_tuned_model_output_1/checkpoint-20000/rng_state.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:a347bb6033226e248954f820c00a1fa9b97f433e6a4c98c5ec3c120eaa8237e2
|
| 3 |
+
size 14244
|
fine_tuned_model_output_1/checkpoint-20000/scaler.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:d57dc504c26fe8dd622f344a68350ece1ae517b34cba92f36da935cbfe157ef0
|
| 3 |
+
size 988
|
fine_tuned_model_output_1/checkpoint-20000/scheduler.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:ca755a82fca1abac6170e402625bdddfb79396e6fed401a1eda74c8a97f738e3
|
| 3 |
+
size 1064
|
fine_tuned_model_output_1/checkpoint-20000/special_tokens_map.json
ADDED
|
@@ -0,0 +1,30 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"bos_token": {
|
| 3 |
+
"content": "<s>",
|
| 4 |
+
"lstrip": false,
|
| 5 |
+
"normalized": false,
|
| 6 |
+
"rstrip": false,
|
| 7 |
+
"single_word": false
|
| 8 |
+
},
|
| 9 |
+
"eos_token": {
|
| 10 |
+
"content": "</s>",
|
| 11 |
+
"lstrip": false,
|
| 12 |
+
"normalized": false,
|
| 13 |
+
"rstrip": false,
|
| 14 |
+
"single_word": false
|
| 15 |
+
},
|
| 16 |
+
"pad_token": {
|
| 17 |
+
"content": "</s>",
|
| 18 |
+
"lstrip": false,
|
| 19 |
+
"normalized": false,
|
| 20 |
+
"rstrip": false,
|
| 21 |
+
"single_word": false
|
| 22 |
+
},
|
| 23 |
+
"unk_token": {
|
| 24 |
+
"content": "<unk>",
|
| 25 |
+
"lstrip": false,
|
| 26 |
+
"normalized": false,
|
| 27 |
+
"rstrip": false,
|
| 28 |
+
"single_word": false
|
| 29 |
+
}
|
| 30 |
+
}
|
fine_tuned_model_output_1/checkpoint-20000/tokenizer.json
ADDED
|
The diff for this file is too large to render.
See raw diff
|
|
|
fine_tuned_model_output_1/checkpoint-20000/tokenizer_config.json
ADDED
|
@@ -0,0 +1,67 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"added_tokens_decoder": {
|
| 3 |
+
"0": {
|
| 4 |
+
"content": "<unk>",
|
| 5 |
+
"lstrip": false,
|
| 6 |
+
"normalized": false,
|
| 7 |
+
"rstrip": false,
|
| 8 |
+
"single_word": false,
|
| 9 |
+
"special": true
|
| 10 |
+
},
|
| 11 |
+
"1": {
|
| 12 |
+
"content": "<s>",
|
| 13 |
+
"lstrip": false,
|
| 14 |
+
"normalized": false,
|
| 15 |
+
"rstrip": false,
|
| 16 |
+
"single_word": false,
|
| 17 |
+
"special": true
|
| 18 |
+
},
|
| 19 |
+
"2": {
|
| 20 |
+
"content": "</s>",
|
| 21 |
+
"lstrip": false,
|
| 22 |
+
"normalized": false,
|
| 23 |
+
"rstrip": false,
|
| 24 |
+
"single_word": false,
|
| 25 |
+
"special": true
|
| 26 |
+
},
|
| 27 |
+
"3": {
|
| 28 |
+
"content": "<pad>",
|
| 29 |
+
"lstrip": false,
|
| 30 |
+
"normalized": false,
|
| 31 |
+
"rstrip": false,
|
| 32 |
+
"single_word": false,
|
| 33 |
+
"special": true
|
| 34 |
+
},
|
| 35 |
+
"32000": {
|
| 36 |
+
"content": "Sanskrit:",
|
| 37 |
+
"lstrip": false,
|
| 38 |
+
"normalized": true,
|
| 39 |
+
"rstrip": false,
|
| 40 |
+
"single_word": false,
|
| 41 |
+
"special": false
|
| 42 |
+
},
|
| 43 |
+
"32001": {
|
| 44 |
+
"content": "English:",
|
| 45 |
+
"lstrip": false,
|
| 46 |
+
"normalized": true,
|
| 47 |
+
"rstrip": false,
|
| 48 |
+
"single_word": false,
|
| 49 |
+
"special": false
|
| 50 |
+
}
|
| 51 |
+
},
|
| 52 |
+
"bos_token": "<s>",
|
| 53 |
+
"clean_up_tokenization_spaces": false,
|
| 54 |
+
"eos_token": "</s>",
|
| 55 |
+
"extra_special_tokens": {},
|
| 56 |
+
"max_length": 512,
|
| 57 |
+
"model_max_length": 1000000000000000019884624838656,
|
| 58 |
+
"pad_to_multiple_of": null,
|
| 59 |
+
"pad_token": "</s>",
|
| 60 |
+
"pad_token_type_id": 0,
|
| 61 |
+
"padding_side": "right",
|
| 62 |
+
"stride": 0,
|
| 63 |
+
"tokenizer_class": "PreTrainedTokenizer",
|
| 64 |
+
"truncation_side": "right",
|
| 65 |
+
"truncation_strategy": "longest_first",
|
| 66 |
+
"unk_token": "<unk>"
|
| 67 |
+
}
|
fine_tuned_model_output_1/checkpoint-20000/trainer_state.json
ADDED
|
@@ -0,0 +1,1593 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": 4.4285101890563965,
|
| 3 |
+
"best_model_checkpoint": "./fine_tuned_model_output_1\\checkpoint-20000",
|
| 4 |
+
"epoch": 0.6319195187300946,
|
| 5 |
+
"eval_steps": 1000,
|
| 6 |
+
"global_step": 20000,
|
| 7 |
+
"is_hyper_param_search": false,
|
| 8 |
+
"is_local_process_zero": true,
|
| 9 |
+
"is_world_process_zero": true,
|
| 10 |
+
"log_history": [
|
| 11 |
+
{
|
| 12 |
+
"epoch": 0.0031595975936504726,
|
| 13 |
+
"grad_norm": 5.00691556930542,
|
| 14 |
+
"learning_rate": 4.984517678283674e-05,
|
| 15 |
+
"loss": 5.1637,
|
| 16 |
+
"step": 100
|
| 17 |
+
},
|
| 18 |
+
{
|
| 19 |
+
"epoch": 0.006319195187300945,
|
| 20 |
+
"grad_norm": 4.790140628814697,
|
| 21 |
+
"learning_rate": 4.968719390818036e-05,
|
| 22 |
+
"loss": 4.9551,
|
| 23 |
+
"step": 200
|
| 24 |
+
},
|
| 25 |
+
{
|
| 26 |
+
"epoch": 0.009478792780951418,
|
| 27 |
+
"grad_norm": 4.75383186340332,
|
| 28 |
+
"learning_rate": 4.952921103352397e-05,
|
| 29 |
+
"loss": 4.8786,
|
| 30 |
+
"step": 300
|
| 31 |
+
},
|
| 32 |
+
{
|
| 33 |
+
"epoch": 0.01263839037460189,
|
| 34 |
+
"grad_norm": 5.218325138092041,
|
| 35 |
+
"learning_rate": 4.9371228158867584e-05,
|
| 36 |
+
"loss": 4.8117,
|
| 37 |
+
"step": 400
|
| 38 |
+
},
|
| 39 |
+
{
|
| 40 |
+
"epoch": 0.015797987968252363,
|
| 41 |
+
"grad_norm": 5.101843357086182,
|
| 42 |
+
"learning_rate": 4.9213245284211194e-05,
|
| 43 |
+
"loss": 4.8153,
|
| 44 |
+
"step": 500
|
| 45 |
+
},
|
| 46 |
+
{
|
| 47 |
+
"epoch": 0.018957585561902836,
|
| 48 |
+
"grad_norm": 4.165743350982666,
|
| 49 |
+
"learning_rate": 4.905526240955481e-05,
|
| 50 |
+
"loss": 4.7331,
|
| 51 |
+
"step": 600
|
| 52 |
+
},
|
| 53 |
+
{
|
| 54 |
+
"epoch": 0.02211718315555331,
|
| 55 |
+
"grad_norm": 4.386114597320557,
|
| 56 |
+
"learning_rate": 4.889727953489842e-05,
|
| 57 |
+
"loss": 4.7982,
|
| 58 |
+
"step": 700
|
| 59 |
+
},
|
| 60 |
+
{
|
| 61 |
+
"epoch": 0.02527678074920378,
|
| 62 |
+
"grad_norm": 4.677374362945557,
|
| 63 |
+
"learning_rate": 4.8739296660242035e-05,
|
| 64 |
+
"loss": 4.751,
|
| 65 |
+
"step": 800
|
| 66 |
+
},
|
| 67 |
+
{
|
| 68 |
+
"epoch": 0.028436378342854254,
|
| 69 |
+
"grad_norm": 4.1365814208984375,
|
| 70 |
+
"learning_rate": 4.8581313785585644e-05,
|
| 71 |
+
"loss": 4.7631,
|
| 72 |
+
"step": 900
|
| 73 |
+
},
|
| 74 |
+
{
|
| 75 |
+
"epoch": 0.031595975936504726,
|
| 76 |
+
"grad_norm": 4.651081562042236,
|
| 77 |
+
"learning_rate": 4.842333091092926e-05,
|
| 78 |
+
"loss": 4.7504,
|
| 79 |
+
"step": 1000
|
| 80 |
+
},
|
| 81 |
+
{
|
| 82 |
+
"epoch": 0.031595975936504726,
|
| 83 |
+
"eval_loss": 4.873119354248047,
|
| 84 |
+
"eval_runtime": 665.3679,
|
| 85 |
+
"eval_samples_per_second": 132.131,
|
| 86 |
+
"eval_steps_per_second": 16.517,
|
| 87 |
+
"step": 1000
|
| 88 |
+
},
|
| 89 |
+
{
|
| 90 |
+
"epoch": 0.0347555735301552,
|
| 91 |
+
"grad_norm": 5.823278903961182,
|
| 92 |
+
"learning_rate": 4.826534803627287e-05,
|
| 93 |
+
"loss": 4.6999,
|
| 94 |
+
"step": 1100
|
| 95 |
+
},
|
| 96 |
+
{
|
| 97 |
+
"epoch": 0.03791517112380567,
|
| 98 |
+
"grad_norm": 4.6261773109436035,
|
| 99 |
+
"learning_rate": 4.8107365161616486e-05,
|
| 100 |
+
"loss": 4.6704,
|
| 101 |
+
"step": 1200
|
| 102 |
+
},
|
| 103 |
+
{
|
| 104 |
+
"epoch": 0.04107476871745615,
|
| 105 |
+
"grad_norm": 4.6412200927734375,
|
| 106 |
+
"learning_rate": 4.79493822869601e-05,
|
| 107 |
+
"loss": 4.691,
|
| 108 |
+
"step": 1300
|
| 109 |
+
},
|
| 110 |
+
{
|
| 111 |
+
"epoch": 0.04423436631110662,
|
| 112 |
+
"grad_norm": 4.368412494659424,
|
| 113 |
+
"learning_rate": 4.779139941230371e-05,
|
| 114 |
+
"loss": 4.7048,
|
| 115 |
+
"step": 1400
|
| 116 |
+
},
|
| 117 |
+
{
|
| 118 |
+
"epoch": 0.04739396390475709,
|
| 119 |
+
"grad_norm": 3.9410157203674316,
|
| 120 |
+
"learning_rate": 4.763499636639389e-05,
|
| 121 |
+
"loss": 4.6213,
|
| 122 |
+
"step": 1500
|
| 123 |
+
},
|
| 124 |
+
{
|
| 125 |
+
"epoch": 0.05055356149840756,
|
| 126 |
+
"grad_norm": 5.697903633117676,
|
| 127 |
+
"learning_rate": 4.74770134917375e-05,
|
| 128 |
+
"loss": 4.6791,
|
| 129 |
+
"step": 1600
|
| 130 |
+
},
|
| 131 |
+
{
|
| 132 |
+
"epoch": 0.05371315909205804,
|
| 133 |
+
"grad_norm": 4.182404041290283,
|
| 134 |
+
"learning_rate": 4.7319030617081114e-05,
|
| 135 |
+
"loss": 4.6493,
|
| 136 |
+
"step": 1700
|
| 137 |
+
},
|
| 138 |
+
{
|
| 139 |
+
"epoch": 0.05687275668570851,
|
| 140 |
+
"grad_norm": 3.8856332302093506,
|
| 141 |
+
"learning_rate": 4.7161047742424724e-05,
|
| 142 |
+
"loss": 4.6081,
|
| 143 |
+
"step": 1800
|
| 144 |
+
},
|
| 145 |
+
{
|
| 146 |
+
"epoch": 0.06003235427935898,
|
| 147 |
+
"grad_norm": 3.778503656387329,
|
| 148 |
+
"learning_rate": 4.700306486776834e-05,
|
| 149 |
+
"loss": 4.6437,
|
| 150 |
+
"step": 1900
|
| 151 |
+
},
|
| 152 |
+
{
|
| 153 |
+
"epoch": 0.06319195187300945,
|
| 154 |
+
"grad_norm": 5.072572231292725,
|
| 155 |
+
"learning_rate": 4.684508199311195e-05,
|
| 156 |
+
"loss": 4.5587,
|
| 157 |
+
"step": 2000
|
| 158 |
+
},
|
| 159 |
+
{
|
| 160 |
+
"epoch": 0.06319195187300945,
|
| 161 |
+
"eval_loss": 4.763667106628418,
|
| 162 |
+
"eval_runtime": 665.5888,
|
| 163 |
+
"eval_samples_per_second": 132.088,
|
| 164 |
+
"eval_steps_per_second": 16.512,
|
| 165 |
+
"step": 2000
|
| 166 |
+
},
|
| 167 |
+
{
|
| 168 |
+
"epoch": 0.06635154946665993,
|
| 169 |
+
"grad_norm": 5.395784854888916,
|
| 170 |
+
"learning_rate": 4.6687099118455565e-05,
|
| 171 |
+
"loss": 4.5621,
|
| 172 |
+
"step": 2100
|
| 173 |
+
},
|
| 174 |
+
{
|
| 175 |
+
"epoch": 0.0695111470603104,
|
| 176 |
+
"grad_norm": 4.856666088104248,
|
| 177 |
+
"learning_rate": 4.6529116243799174e-05,
|
| 178 |
+
"loss": 4.5721,
|
| 179 |
+
"step": 2200
|
| 180 |
+
},
|
| 181 |
+
{
|
| 182 |
+
"epoch": 0.07267074465396087,
|
| 183 |
+
"grad_norm": 4.935929298400879,
|
| 184 |
+
"learning_rate": 4.637113336914279e-05,
|
| 185 |
+
"loss": 4.6004,
|
| 186 |
+
"step": 2300
|
| 187 |
+
},
|
| 188 |
+
{
|
| 189 |
+
"epoch": 0.07583034224761134,
|
| 190 |
+
"grad_norm": 5.257358551025391,
|
| 191 |
+
"learning_rate": 4.62131504944864e-05,
|
| 192 |
+
"loss": 4.6032,
|
| 193 |
+
"step": 2400
|
| 194 |
+
},
|
| 195 |
+
{
|
| 196 |
+
"epoch": 0.07898993984126182,
|
| 197 |
+
"grad_norm": 4.67532205581665,
|
| 198 |
+
"learning_rate": 4.6055167619830016e-05,
|
| 199 |
+
"loss": 4.5734,
|
| 200 |
+
"step": 2500
|
| 201 |
+
},
|
| 202 |
+
{
|
| 203 |
+
"epoch": 0.0821495374349123,
|
| 204 |
+
"grad_norm": 4.023921966552734,
|
| 205 |
+
"learning_rate": 4.5897184745173625e-05,
|
| 206 |
+
"loss": 4.5888,
|
| 207 |
+
"step": 2600
|
| 208 |
+
},
|
| 209 |
+
{
|
| 210 |
+
"epoch": 0.08530913502856276,
|
| 211 |
+
"grad_norm": 3.9619414806365967,
|
| 212 |
+
"learning_rate": 4.573920187051724e-05,
|
| 213 |
+
"loss": 4.5703,
|
| 214 |
+
"step": 2700
|
| 215 |
+
},
|
| 216 |
+
{
|
| 217 |
+
"epoch": 0.08846873262221323,
|
| 218 |
+
"grad_norm": 3.7340047359466553,
|
| 219 |
+
"learning_rate": 4.558121899586085e-05,
|
| 220 |
+
"loss": 4.5437,
|
| 221 |
+
"step": 2800
|
| 222 |
+
},
|
| 223 |
+
{
|
| 224 |
+
"epoch": 0.09162833021586371,
|
| 225 |
+
"grad_norm": 3.5358519554138184,
|
| 226 |
+
"learning_rate": 4.5423236121204466e-05,
|
| 227 |
+
"loss": 4.5433,
|
| 228 |
+
"step": 2900
|
| 229 |
+
},
|
| 230 |
+
{
|
| 231 |
+
"epoch": 0.09478792780951419,
|
| 232 |
+
"grad_norm": 5.275386810302734,
|
| 233 |
+
"learning_rate": 4.5265253246548075e-05,
|
| 234 |
+
"loss": 4.5315,
|
| 235 |
+
"step": 3000
|
| 236 |
+
},
|
| 237 |
+
{
|
| 238 |
+
"epoch": 0.09478792780951419,
|
| 239 |
+
"eval_loss": 4.6958231925964355,
|
| 240 |
+
"eval_runtime": 660.1374,
|
| 241 |
+
"eval_samples_per_second": 133.178,
|
| 242 |
+
"eval_steps_per_second": 16.648,
|
| 243 |
+
"step": 3000
|
| 244 |
+
},
|
| 245 |
+
{
|
| 246 |
+
"epoch": 0.09794752540316465,
|
| 247 |
+
"grad_norm": 4.195945739746094,
|
| 248 |
+
"learning_rate": 4.510727037189169e-05,
|
| 249 |
+
"loss": 4.5292,
|
| 250 |
+
"step": 3100
|
| 251 |
+
},
|
| 252 |
+
{
|
| 253 |
+
"epoch": 0.10110712299681512,
|
| 254 |
+
"grad_norm": 4.976147651672363,
|
| 255 |
+
"learning_rate": 4.49492874972353e-05,
|
| 256 |
+
"loss": 4.5718,
|
| 257 |
+
"step": 3200
|
| 258 |
+
},
|
| 259 |
+
{
|
| 260 |
+
"epoch": 0.1042667205904656,
|
| 261 |
+
"grad_norm": 3.999481678009033,
|
| 262 |
+
"learning_rate": 4.479130462257892e-05,
|
| 263 |
+
"loss": 4.5177,
|
| 264 |
+
"step": 3300
|
| 265 |
+
},
|
| 266 |
+
{
|
| 267 |
+
"epoch": 0.10742631818411608,
|
| 268 |
+
"grad_norm": 5.00895357131958,
|
| 269 |
+
"learning_rate": 4.463332174792253e-05,
|
| 270 |
+
"loss": 4.5418,
|
| 271 |
+
"step": 3400
|
| 272 |
+
},
|
| 273 |
+
{
|
| 274 |
+
"epoch": 0.11058591577776654,
|
| 275 |
+
"grad_norm": 4.499265193939209,
|
| 276 |
+
"learning_rate": 4.447533887326614e-05,
|
| 277 |
+
"loss": 4.5262,
|
| 278 |
+
"step": 3500
|
| 279 |
+
},
|
| 280 |
+
{
|
| 281 |
+
"epoch": 0.11374551337141701,
|
| 282 |
+
"grad_norm": 4.686629772186279,
|
| 283 |
+
"learning_rate": 4.431735599860976e-05,
|
| 284 |
+
"loss": 4.5276,
|
| 285 |
+
"step": 3600
|
| 286 |
+
},
|
| 287 |
+
{
|
| 288 |
+
"epoch": 0.11690511096506749,
|
| 289 |
+
"grad_norm": 5.327017784118652,
|
| 290 |
+
"learning_rate": 4.415937312395337e-05,
|
| 291 |
+
"loss": 4.4939,
|
| 292 |
+
"step": 3700
|
| 293 |
+
},
|
| 294 |
+
{
|
| 295 |
+
"epoch": 0.12006470855871797,
|
| 296 |
+
"grad_norm": 3.8358302116394043,
|
| 297 |
+
"learning_rate": 4.400139024929698e-05,
|
| 298 |
+
"loss": 4.5028,
|
| 299 |
+
"step": 3800
|
| 300 |
+
},
|
| 301 |
+
{
|
| 302 |
+
"epoch": 0.12322430615236843,
|
| 303 |
+
"grad_norm": 4.16530704498291,
|
| 304 |
+
"learning_rate": 4.384340737464059e-05,
|
| 305 |
+
"loss": 4.4964,
|
| 306 |
+
"step": 3900
|
| 307 |
+
},
|
| 308 |
+
{
|
| 309 |
+
"epoch": 0.1263839037460189,
|
| 310 |
+
"grad_norm": 4.142209053039551,
|
| 311 |
+
"learning_rate": 4.36854244999842e-05,
|
| 312 |
+
"loss": 4.4899,
|
| 313 |
+
"step": 4000
|
| 314 |
+
},
|
| 315 |
+
{
|
| 316 |
+
"epoch": 0.1263839037460189,
|
| 317 |
+
"eval_loss": 4.649657249450684,
|
| 318 |
+
"eval_runtime": 689.9636,
|
| 319 |
+
"eval_samples_per_second": 127.421,
|
| 320 |
+
"eval_steps_per_second": 15.928,
|
| 321 |
+
"step": 4000
|
| 322 |
+
},
|
| 323 |
+
{
|
| 324 |
+
"epoch": 0.12954350133966938,
|
| 325 |
+
"grad_norm": 3.615328311920166,
|
| 326 |
+
"learning_rate": 4.352744162532781e-05,
|
| 327 |
+
"loss": 4.4671,
|
| 328 |
+
"step": 4100
|
| 329 |
+
},
|
| 330 |
+
{
|
| 331 |
+
"epoch": 0.13270309893331986,
|
| 332 |
+
"grad_norm": 5.139493465423584,
|
| 333 |
+
"learning_rate": 4.336945875067143e-05,
|
| 334 |
+
"loss": 4.4954,
|
| 335 |
+
"step": 4200
|
| 336 |
+
},
|
| 337 |
+
{
|
| 338 |
+
"epoch": 0.13586269652697033,
|
| 339 |
+
"grad_norm": 4.381932258605957,
|
| 340 |
+
"learning_rate": 4.3211475876015036e-05,
|
| 341 |
+
"loss": 4.5103,
|
| 342 |
+
"step": 4300
|
| 343 |
+
},
|
| 344 |
+
{
|
| 345 |
+
"epoch": 0.1390222941206208,
|
| 346 |
+
"grad_norm": 4.257575988769531,
|
| 347 |
+
"learning_rate": 4.305349300135865e-05,
|
| 348 |
+
"loss": 4.4677,
|
| 349 |
+
"step": 4400
|
| 350 |
+
},
|
| 351 |
+
{
|
| 352 |
+
"epoch": 0.14218189171427126,
|
| 353 |
+
"grad_norm": 3.5805318355560303,
|
| 354 |
+
"learning_rate": 4.289551012670227e-05,
|
| 355 |
+
"loss": 4.4547,
|
| 356 |
+
"step": 4500
|
| 357 |
+
},
|
| 358 |
+
{
|
| 359 |
+
"epoch": 0.14534148930792173,
|
| 360 |
+
"grad_norm": 4.602927207946777,
|
| 361 |
+
"learning_rate": 4.273752725204588e-05,
|
| 362 |
+
"loss": 4.4885,
|
| 363 |
+
"step": 4600
|
| 364 |
+
},
|
| 365 |
+
{
|
| 366 |
+
"epoch": 0.1485010869015722,
|
| 367 |
+
"grad_norm": 4.054294586181641,
|
| 368 |
+
"learning_rate": 4.2579544377389494e-05,
|
| 369 |
+
"loss": 4.4791,
|
| 370 |
+
"step": 4700
|
| 371 |
+
},
|
| 372 |
+
{
|
| 373 |
+
"epoch": 0.1516606844952227,
|
| 374 |
+
"grad_norm": 3.857389450073242,
|
| 375 |
+
"learning_rate": 4.24215615027331e-05,
|
| 376 |
+
"loss": 4.4316,
|
| 377 |
+
"step": 4800
|
| 378 |
+
},
|
| 379 |
+
{
|
| 380 |
+
"epoch": 0.15482028208887316,
|
| 381 |
+
"grad_norm": 3.913726806640625,
|
| 382 |
+
"learning_rate": 4.226357862807672e-05,
|
| 383 |
+
"loss": 4.4786,
|
| 384 |
+
"step": 4900
|
| 385 |
+
},
|
| 386 |
+
{
|
| 387 |
+
"epoch": 0.15797987968252364,
|
| 388 |
+
"grad_norm": 4.159733295440674,
|
| 389 |
+
"learning_rate": 4.210559575342033e-05,
|
| 390 |
+
"loss": 4.422,
|
| 391 |
+
"step": 5000
|
| 392 |
+
},
|
| 393 |
+
{
|
| 394 |
+
"epoch": 0.15797987968252364,
|
| 395 |
+
"eval_loss": 4.6132493019104,
|
| 396 |
+
"eval_runtime": 721.543,
|
| 397 |
+
"eval_samples_per_second": 121.844,
|
| 398 |
+
"eval_steps_per_second": 15.231,
|
| 399 |
+
"step": 5000
|
| 400 |
+
},
|
| 401 |
+
{
|
| 402 |
+
"epoch": 0.16113947727617411,
|
| 403 |
+
"grad_norm": 5.403946876525879,
|
| 404 |
+
"learning_rate": 4.1947612878763944e-05,
|
| 405 |
+
"loss": 4.4857,
|
| 406 |
+
"step": 5100
|
| 407 |
+
},
|
| 408 |
+
{
|
| 409 |
+
"epoch": 0.1642990748698246,
|
| 410 |
+
"grad_norm": 11.68526554107666,
|
| 411 |
+
"learning_rate": 4.178963000410755e-05,
|
| 412 |
+
"loss": 4.4777,
|
| 413 |
+
"step": 5200
|
| 414 |
+
},
|
| 415 |
+
{
|
| 416 |
+
"epoch": 0.16745867246347504,
|
| 417 |
+
"grad_norm": 4.937869548797607,
|
| 418 |
+
"learning_rate": 4.163164712945117e-05,
|
| 419 |
+
"loss": 4.4452,
|
| 420 |
+
"step": 5300
|
| 421 |
+
},
|
| 422 |
+
{
|
| 423 |
+
"epoch": 0.17061827005712552,
|
| 424 |
+
"grad_norm": 3.7432169914245605,
|
| 425 |
+
"learning_rate": 4.147366425479478e-05,
|
| 426 |
+
"loss": 4.4717,
|
| 427 |
+
"step": 5400
|
| 428 |
+
},
|
| 429 |
+
{
|
| 430 |
+
"epoch": 0.173777867650776,
|
| 431 |
+
"grad_norm": 3.801788806915283,
|
| 432 |
+
"learning_rate": 4.1317261208884964e-05,
|
| 433 |
+
"loss": 4.4424,
|
| 434 |
+
"step": 5500
|
| 435 |
+
},
|
| 436 |
+
{
|
| 437 |
+
"epoch": 0.17693746524442647,
|
| 438 |
+
"grad_norm": 4.984846115112305,
|
| 439 |
+
"learning_rate": 4.115927833422857e-05,
|
| 440 |
+
"loss": 4.4671,
|
| 441 |
+
"step": 5600
|
| 442 |
+
},
|
| 443 |
+
{
|
| 444 |
+
"epoch": 0.18009706283807694,
|
| 445 |
+
"grad_norm": 3.8125479221343994,
|
| 446 |
+
"learning_rate": 4.100287528831875e-05,
|
| 447 |
+
"loss": 4.4079,
|
| 448 |
+
"step": 5700
|
| 449 |
+
},
|
| 450 |
+
{
|
| 451 |
+
"epoch": 0.18325666043172742,
|
| 452 |
+
"grad_norm": 5.528467655181885,
|
| 453 |
+
"learning_rate": 4.084489241366236e-05,
|
| 454 |
+
"loss": 4.4194,
|
| 455 |
+
"step": 5800
|
| 456 |
+
},
|
| 457 |
+
{
|
| 458 |
+
"epoch": 0.1864162580253779,
|
| 459 |
+
"grad_norm": 4.0301899909973145,
|
| 460 |
+
"learning_rate": 4.068690953900598e-05,
|
| 461 |
+
"loss": 4.4299,
|
| 462 |
+
"step": 5900
|
| 463 |
+
},
|
| 464 |
+
{
|
| 465 |
+
"epoch": 0.18957585561902837,
|
| 466 |
+
"grad_norm": 3.4992246627807617,
|
| 467 |
+
"learning_rate": 4.0528926664349586e-05,
|
| 468 |
+
"loss": 4.448,
|
| 469 |
+
"step": 6000
|
| 470 |
+
},
|
| 471 |
+
{
|
| 472 |
+
"epoch": 0.18957585561902837,
|
| 473 |
+
"eval_loss": 4.5857086181640625,
|
| 474 |
+
"eval_runtime": 702.1155,
|
| 475 |
+
"eval_samples_per_second": 125.216,
|
| 476 |
+
"eval_steps_per_second": 15.653,
|
| 477 |
+
"step": 6000
|
| 478 |
+
},
|
| 479 |
+
{
|
| 480 |
+
"epoch": 0.19273545321267882,
|
| 481 |
+
"grad_norm": 4.121975421905518,
|
| 482 |
+
"learning_rate": 4.03709437896932e-05,
|
| 483 |
+
"loss": 4.4076,
|
| 484 |
+
"step": 6100
|
| 485 |
+
},
|
| 486 |
+
{
|
| 487 |
+
"epoch": 0.1958950508063293,
|
| 488 |
+
"grad_norm": 3.662929058074951,
|
| 489 |
+
"learning_rate": 4.021296091503681e-05,
|
| 490 |
+
"loss": 4.3953,
|
| 491 |
+
"step": 6200
|
| 492 |
+
},
|
| 493 |
+
{
|
| 494 |
+
"epoch": 0.19905464839997977,
|
| 495 |
+
"grad_norm": 4.008944988250732,
|
| 496 |
+
"learning_rate": 4.005497804038043e-05,
|
| 497 |
+
"loss": 4.4581,
|
| 498 |
+
"step": 6300
|
| 499 |
+
},
|
| 500 |
+
{
|
| 501 |
+
"epoch": 0.20221424599363025,
|
| 502 |
+
"grad_norm": 3.7961764335632324,
|
| 503 |
+
"learning_rate": 3.9896995165724036e-05,
|
| 504 |
+
"loss": 4.397,
|
| 505 |
+
"step": 6400
|
| 506 |
+
},
|
| 507 |
+
{
|
| 508 |
+
"epoch": 0.20537384358728072,
|
| 509 |
+
"grad_norm": 4.710026741027832,
|
| 510 |
+
"learning_rate": 3.973901229106765e-05,
|
| 511 |
+
"loss": 4.4515,
|
| 512 |
+
"step": 6500
|
| 513 |
+
},
|
| 514 |
+
{
|
| 515 |
+
"epoch": 0.2085334411809312,
|
| 516 |
+
"grad_norm": 4.220251083374023,
|
| 517 |
+
"learning_rate": 3.958102941641126e-05,
|
| 518 |
+
"loss": 4.4199,
|
| 519 |
+
"step": 6600
|
| 520 |
+
},
|
| 521 |
+
{
|
| 522 |
+
"epoch": 0.21169303877458168,
|
| 523 |
+
"grad_norm": 3.7243406772613525,
|
| 524 |
+
"learning_rate": 3.942304654175488e-05,
|
| 525 |
+
"loss": 4.4097,
|
| 526 |
+
"step": 6700
|
| 527 |
+
},
|
| 528 |
+
{
|
| 529 |
+
"epoch": 0.21485263636823215,
|
| 530 |
+
"grad_norm": 4.340923309326172,
|
| 531 |
+
"learning_rate": 3.926664349584505e-05,
|
| 532 |
+
"loss": 4.4105,
|
| 533 |
+
"step": 6800
|
| 534 |
+
},
|
| 535 |
+
{
|
| 536 |
+
"epoch": 0.2180122339618826,
|
| 537 |
+
"grad_norm": 4.304684638977051,
|
| 538 |
+
"learning_rate": 3.9108660621188665e-05,
|
| 539 |
+
"loss": 4.4033,
|
| 540 |
+
"step": 6900
|
| 541 |
+
},
|
| 542 |
+
{
|
| 543 |
+
"epoch": 0.22117183155553308,
|
| 544 |
+
"grad_norm": 4.587896347045898,
|
| 545 |
+
"learning_rate": 3.8950677746532275e-05,
|
| 546 |
+
"loss": 4.4451,
|
| 547 |
+
"step": 7000
|
| 548 |
+
},
|
| 549 |
+
{
|
| 550 |
+
"epoch": 0.22117183155553308,
|
| 551 |
+
"eval_loss": 4.5615153312683105,
|
| 552 |
+
"eval_runtime": 695.974,
|
| 553 |
+
"eval_samples_per_second": 126.321,
|
| 554 |
+
"eval_steps_per_second": 15.791,
|
| 555 |
+
"step": 7000
|
| 556 |
+
},
|
| 557 |
+
{
|
| 558 |
+
"epoch": 0.22433142914918355,
|
| 559 |
+
"grad_norm": 4.658904075622559,
|
| 560 |
+
"learning_rate": 3.879269487187589e-05,
|
| 561 |
+
"loss": 4.407,
|
| 562 |
+
"step": 7100
|
| 563 |
+
},
|
| 564 |
+
{
|
| 565 |
+
"epoch": 0.22749102674283403,
|
| 566 |
+
"grad_norm": 4.226349830627441,
|
| 567 |
+
"learning_rate": 3.86347119972195e-05,
|
| 568 |
+
"loss": 4.4055,
|
| 569 |
+
"step": 7200
|
| 570 |
+
},
|
| 571 |
+
{
|
| 572 |
+
"epoch": 0.2306506243364845,
|
| 573 |
+
"grad_norm": 5.705229759216309,
|
| 574 |
+
"learning_rate": 3.8476729122563116e-05,
|
| 575 |
+
"loss": 4.3862,
|
| 576 |
+
"step": 7300
|
| 577 |
+
},
|
| 578 |
+
{
|
| 579 |
+
"epoch": 0.23381022193013498,
|
| 580 |
+
"grad_norm": 3.9644031524658203,
|
| 581 |
+
"learning_rate": 3.8318746247906725e-05,
|
| 582 |
+
"loss": 4.3647,
|
| 583 |
+
"step": 7400
|
| 584 |
+
},
|
| 585 |
+
{
|
| 586 |
+
"epoch": 0.23696981952378546,
|
| 587 |
+
"grad_norm": 4.256899356842041,
|
| 588 |
+
"learning_rate": 3.816076337325034e-05,
|
| 589 |
+
"loss": 4.4091,
|
| 590 |
+
"step": 7500
|
| 591 |
+
},
|
| 592 |
+
{
|
| 593 |
+
"epoch": 0.24012941711743593,
|
| 594 |
+
"grad_norm": 3.951305627822876,
|
| 595 |
+
"learning_rate": 3.800278049859396e-05,
|
| 596 |
+
"loss": 4.3994,
|
| 597 |
+
"step": 7600
|
| 598 |
+
},
|
| 599 |
+
{
|
| 600 |
+
"epoch": 0.2432890147110864,
|
| 601 |
+
"grad_norm": 3.6965837478637695,
|
| 602 |
+
"learning_rate": 3.7844797623937566e-05,
|
| 603 |
+
"loss": 4.4043,
|
| 604 |
+
"step": 7700
|
| 605 |
+
},
|
| 606 |
+
{
|
| 607 |
+
"epoch": 0.24644861230473686,
|
| 608 |
+
"grad_norm": 4.412446022033691,
|
| 609 |
+
"learning_rate": 3.768681474928118e-05,
|
| 610 |
+
"loss": 4.4237,
|
| 611 |
+
"step": 7800
|
| 612 |
+
},
|
| 613 |
+
{
|
| 614 |
+
"epoch": 0.24960820989838733,
|
| 615 |
+
"grad_norm": 4.190601348876953,
|
| 616 |
+
"learning_rate": 3.752883187462479e-05,
|
| 617 |
+
"loss": 4.365,
|
| 618 |
+
"step": 7900
|
| 619 |
+
},
|
| 620 |
+
{
|
| 621 |
+
"epoch": 0.2527678074920378,
|
| 622 |
+
"grad_norm": 3.6600217819213867,
|
| 623 |
+
"learning_rate": 3.737084899996841e-05,
|
| 624 |
+
"loss": 4.393,
|
| 625 |
+
"step": 8000
|
| 626 |
+
},
|
| 627 |
+
{
|
| 628 |
+
"epoch": 0.2527678074920378,
|
| 629 |
+
"eval_loss": 4.542718887329102,
|
| 630 |
+
"eval_runtime": 692.645,
|
| 631 |
+
"eval_samples_per_second": 126.928,
|
| 632 |
+
"eval_steps_per_second": 15.867,
|
| 633 |
+
"step": 8000
|
| 634 |
+
},
|
| 635 |
+
{
|
| 636 |
+
"epoch": 0.2559274050856883,
|
| 637 |
+
"grad_norm": 6.242320537567139,
|
| 638 |
+
"learning_rate": 3.721286612531202e-05,
|
| 639 |
+
"loss": 4.4185,
|
| 640 |
+
"step": 8100
|
| 641 |
+
},
|
| 642 |
+
{
|
| 643 |
+
"epoch": 0.25908700267933876,
|
| 644 |
+
"grad_norm": 4.762045383453369,
|
| 645 |
+
"learning_rate": 3.705488325065563e-05,
|
| 646 |
+
"loss": 4.3656,
|
| 647 |
+
"step": 8200
|
| 648 |
+
},
|
| 649 |
+
{
|
| 650 |
+
"epoch": 0.26224660027298924,
|
| 651 |
+
"grad_norm": 5.066886901855469,
|
| 652 |
+
"learning_rate": 3.689690037599924e-05,
|
| 653 |
+
"loss": 4.394,
|
| 654 |
+
"step": 8300
|
| 655 |
+
},
|
| 656 |
+
{
|
| 657 |
+
"epoch": 0.2654061978666397,
|
| 658 |
+
"grad_norm": 4.975124835968018,
|
| 659 |
+
"learning_rate": 3.673891750134286e-05,
|
| 660 |
+
"loss": 4.3346,
|
| 661 |
+
"step": 8400
|
| 662 |
+
},
|
| 663 |
+
{
|
| 664 |
+
"epoch": 0.2685657954602902,
|
| 665 |
+
"grad_norm": 4.436885356903076,
|
| 666 |
+
"learning_rate": 3.658093462668647e-05,
|
| 667 |
+
"loss": 4.3958,
|
| 668 |
+
"step": 8500
|
| 669 |
+
},
|
| 670 |
+
{
|
| 671 |
+
"epoch": 0.27172539305394067,
|
| 672 |
+
"grad_norm": 4.618080139160156,
|
| 673 |
+
"learning_rate": 3.6422951752030084e-05,
|
| 674 |
+
"loss": 4.3791,
|
| 675 |
+
"step": 8600
|
| 676 |
+
},
|
| 677 |
+
{
|
| 678 |
+
"epoch": 0.27488499064759114,
|
| 679 |
+
"grad_norm": 4.878961086273193,
|
| 680 |
+
"learning_rate": 3.626496887737369e-05,
|
| 681 |
+
"loss": 4.3738,
|
| 682 |
+
"step": 8700
|
| 683 |
+
},
|
| 684 |
+
{
|
| 685 |
+
"epoch": 0.2780445882412416,
|
| 686 |
+
"grad_norm": 3.7712392807006836,
|
| 687 |
+
"learning_rate": 3.610698600271731e-05,
|
| 688 |
+
"loss": 4.362,
|
| 689 |
+
"step": 8800
|
| 690 |
+
},
|
| 691 |
+
{
|
| 692 |
+
"epoch": 0.2812041858348921,
|
| 693 |
+
"grad_norm": 4.318520545959473,
|
| 694 |
+
"learning_rate": 3.594900312806092e-05,
|
| 695 |
+
"loss": 4.367,
|
| 696 |
+
"step": 8900
|
| 697 |
+
},
|
| 698 |
+
{
|
| 699 |
+
"epoch": 0.2843637834285425,
|
| 700 |
+
"grad_norm": 3.3742518424987793,
|
| 701 |
+
"learning_rate": 3.5791020253404534e-05,
|
| 702 |
+
"loss": 4.3763,
|
| 703 |
+
"step": 9000
|
| 704 |
+
},
|
| 705 |
+
{
|
| 706 |
+
"epoch": 0.2843637834285425,
|
| 707 |
+
"eval_loss": 4.525923728942871,
|
| 708 |
+
"eval_runtime": 692.5212,
|
| 709 |
+
"eval_samples_per_second": 126.951,
|
| 710 |
+
"eval_steps_per_second": 15.87,
|
| 711 |
+
"step": 9000
|
| 712 |
+
},
|
| 713 |
+
{
|
| 714 |
+
"epoch": 0.287523381022193,
|
| 715 |
+
"grad_norm": 4.027231693267822,
|
| 716 |
+
"learning_rate": 3.5633037378748143e-05,
|
| 717 |
+
"loss": 4.4086,
|
| 718 |
+
"step": 9100
|
| 719 |
+
},
|
| 720 |
+
{
|
| 721 |
+
"epoch": 0.29068297861584347,
|
| 722 |
+
"grad_norm": 4.556132793426514,
|
| 723 |
+
"learning_rate": 3.547505450409176e-05,
|
| 724 |
+
"loss": 4.3686,
|
| 725 |
+
"step": 9200
|
| 726 |
+
},
|
| 727 |
+
{
|
| 728 |
+
"epoch": 0.29384257620949394,
|
| 729 |
+
"grad_norm": 3.9203240871429443,
|
| 730 |
+
"learning_rate": 3.531707162943537e-05,
|
| 731 |
+
"loss": 4.365,
|
| 732 |
+
"step": 9300
|
| 733 |
+
},
|
| 734 |
+
{
|
| 735 |
+
"epoch": 0.2970021738031444,
|
| 736 |
+
"grad_norm": 4.094367027282715,
|
| 737 |
+
"learning_rate": 3.5159088754778985e-05,
|
| 738 |
+
"loss": 4.3734,
|
| 739 |
+
"step": 9400
|
| 740 |
+
},
|
| 741 |
+
{
|
| 742 |
+
"epoch": 0.3001617713967949,
|
| 743 |
+
"grad_norm": 5.018619060516357,
|
| 744 |
+
"learning_rate": 3.50011058801226e-05,
|
| 745 |
+
"loss": 4.4014,
|
| 746 |
+
"step": 9500
|
| 747 |
+
},
|
| 748 |
+
{
|
| 749 |
+
"epoch": 0.3033213689904454,
|
| 750 |
+
"grad_norm": 3.928473711013794,
|
| 751 |
+
"learning_rate": 3.484312300546621e-05,
|
| 752 |
+
"loss": 4.3508,
|
| 753 |
+
"step": 9600
|
| 754 |
+
},
|
| 755 |
+
{
|
| 756 |
+
"epoch": 0.30648096658409585,
|
| 757 |
+
"grad_norm": 3.889299154281616,
|
| 758 |
+
"learning_rate": 3.4685140130809826e-05,
|
| 759 |
+
"loss": 4.365,
|
| 760 |
+
"step": 9700
|
| 761 |
+
},
|
| 762 |
+
{
|
| 763 |
+
"epoch": 0.3096405641777463,
|
| 764 |
+
"grad_norm": 4.932437896728516,
|
| 765 |
+
"learning_rate": 3.4527157256153435e-05,
|
| 766 |
+
"loss": 4.3166,
|
| 767 |
+
"step": 9800
|
| 768 |
+
},
|
| 769 |
+
{
|
| 770 |
+
"epoch": 0.3128001617713968,
|
| 771 |
+
"grad_norm": 4.175967216491699,
|
| 772 |
+
"learning_rate": 3.436917438149705e-05,
|
| 773 |
+
"loss": 4.3463,
|
| 774 |
+
"step": 9900
|
| 775 |
+
},
|
| 776 |
+
{
|
| 777 |
+
"epoch": 0.3159597593650473,
|
| 778 |
+
"grad_norm": 5.54374361038208,
|
| 779 |
+
"learning_rate": 3.421119150684066e-05,
|
| 780 |
+
"loss": 4.3506,
|
| 781 |
+
"step": 10000
|
| 782 |
+
},
|
| 783 |
+
{
|
| 784 |
+
"epoch": 0.3159597593650473,
|
| 785 |
+
"eval_loss": 4.511664390563965,
|
| 786 |
+
"eval_runtime": 691.9773,
|
| 787 |
+
"eval_samples_per_second": 127.05,
|
| 788 |
+
"eval_steps_per_second": 15.882,
|
| 789 |
+
"step": 10000
|
| 790 |
+
},
|
| 791 |
+
{
|
| 792 |
+
"epoch": 0.31911935695869775,
|
| 793 |
+
"grad_norm": 4.466663360595703,
|
| 794 |
+
"learning_rate": 3.4053208632184277e-05,
|
| 795 |
+
"loss": 4.3351,
|
| 796 |
+
"step": 10100
|
| 797 |
+
},
|
| 798 |
+
{
|
| 799 |
+
"epoch": 0.32227895455234823,
|
| 800 |
+
"grad_norm": 4.300159454345703,
|
| 801 |
+
"learning_rate": 3.3895225757527886e-05,
|
| 802 |
+
"loss": 4.3605,
|
| 803 |
+
"step": 10200
|
| 804 |
+
},
|
| 805 |
+
{
|
| 806 |
+
"epoch": 0.3254385521459987,
|
| 807 |
+
"grad_norm": 4.595406532287598,
|
| 808 |
+
"learning_rate": 3.37372428828715e-05,
|
| 809 |
+
"loss": 4.3432,
|
| 810 |
+
"step": 10300
|
| 811 |
+
},
|
| 812 |
+
{
|
| 813 |
+
"epoch": 0.3285981497396492,
|
| 814 |
+
"grad_norm": 4.077473163604736,
|
| 815 |
+
"learning_rate": 3.357926000821511e-05,
|
| 816 |
+
"loss": 4.3421,
|
| 817 |
+
"step": 10400
|
| 818 |
+
},
|
| 819 |
+
{
|
| 820 |
+
"epoch": 0.33175774733329966,
|
| 821 |
+
"grad_norm": 4.250525951385498,
|
| 822 |
+
"learning_rate": 3.342127713355873e-05,
|
| 823 |
+
"loss": 4.3795,
|
| 824 |
+
"step": 10500
|
| 825 |
+
},
|
| 826 |
+
{
|
| 827 |
+
"epoch": 0.3349173449269501,
|
| 828 |
+
"grad_norm": 3.9417407512664795,
|
| 829 |
+
"learning_rate": 3.3263294258902336e-05,
|
| 830 |
+
"loss": 4.3394,
|
| 831 |
+
"step": 10600
|
| 832 |
+
},
|
| 833 |
+
{
|
| 834 |
+
"epoch": 0.33807694252060055,
|
| 835 |
+
"grad_norm": 3.782409906387329,
|
| 836 |
+
"learning_rate": 3.310531138424595e-05,
|
| 837 |
+
"loss": 4.3061,
|
| 838 |
+
"step": 10700
|
| 839 |
+
},
|
| 840 |
+
{
|
| 841 |
+
"epoch": 0.34123654011425103,
|
| 842 |
+
"grad_norm": 4.497501850128174,
|
| 843 |
+
"learning_rate": 3.294732850958956e-05,
|
| 844 |
+
"loss": 4.3558,
|
| 845 |
+
"step": 10800
|
| 846 |
+
},
|
| 847 |
+
{
|
| 848 |
+
"epoch": 0.3443961377079015,
|
| 849 |
+
"grad_norm": 4.938557147979736,
|
| 850 |
+
"learning_rate": 3.278934563493318e-05,
|
| 851 |
+
"loss": 4.33,
|
| 852 |
+
"step": 10900
|
| 853 |
+
},
|
| 854 |
+
{
|
| 855 |
+
"epoch": 0.347555735301552,
|
| 856 |
+
"grad_norm": 4.101445198059082,
|
| 857 |
+
"learning_rate": 3.263136276027679e-05,
|
| 858 |
+
"loss": 4.3658,
|
| 859 |
+
"step": 11000
|
| 860 |
+
},
|
| 861 |
+
{
|
| 862 |
+
"epoch": 0.347555735301552,
|
| 863 |
+
"eval_loss": 4.499057769775391,
|
| 864 |
+
"eval_runtime": 696.9678,
|
| 865 |
+
"eval_samples_per_second": 126.141,
|
| 866 |
+
"eval_steps_per_second": 15.768,
|
| 867 |
+
"step": 11000
|
| 868 |
+
},
|
| 869 |
+
{
|
| 870 |
+
"epoch": 0.35071533289520246,
|
| 871 |
+
"grad_norm": 4.350194454193115,
|
| 872 |
+
"learning_rate": 3.24733798856204e-05,
|
| 873 |
+
"loss": 4.3129,
|
| 874 |
+
"step": 11100
|
| 875 |
+
},
|
| 876 |
+
{
|
| 877 |
+
"epoch": 0.35387493048885293,
|
| 878 |
+
"grad_norm": 4.244603157043457,
|
| 879 |
+
"learning_rate": 3.231539701096401e-05,
|
| 880 |
+
"loss": 4.3566,
|
| 881 |
+
"step": 11200
|
| 882 |
+
},
|
| 883 |
+
{
|
| 884 |
+
"epoch": 0.3570345280825034,
|
| 885 |
+
"grad_norm": 3.947141647338867,
|
| 886 |
+
"learning_rate": 3.215741413630762e-05,
|
| 887 |
+
"loss": 4.3572,
|
| 888 |
+
"step": 11300
|
| 889 |
+
},
|
| 890 |
+
{
|
| 891 |
+
"epoch": 0.3601941256761539,
|
| 892 |
+
"grad_norm": 4.241100311279297,
|
| 893 |
+
"learning_rate": 3.199943126165124e-05,
|
| 894 |
+
"loss": 4.3341,
|
| 895 |
+
"step": 11400
|
| 896 |
+
},
|
| 897 |
+
{
|
| 898 |
+
"epoch": 0.36335372326980436,
|
| 899 |
+
"grad_norm": 4.596353530883789,
|
| 900 |
+
"learning_rate": 3.184144838699485e-05,
|
| 901 |
+
"loss": 4.3702,
|
| 902 |
+
"step": 11500
|
| 903 |
+
},
|
| 904 |
+
{
|
| 905 |
+
"epoch": 0.36651332086345484,
|
| 906 |
+
"grad_norm": 4.667087554931641,
|
| 907 |
+
"learning_rate": 3.168346551233846e-05,
|
| 908 |
+
"loss": 4.3919,
|
| 909 |
+
"step": 11600
|
| 910 |
+
},
|
| 911 |
+
{
|
| 912 |
+
"epoch": 0.3696729184571053,
|
| 913 |
+
"grad_norm": 4.700224876403809,
|
| 914 |
+
"learning_rate": 3.152548263768207e-05,
|
| 915 |
+
"loss": 4.3622,
|
| 916 |
+
"step": 11700
|
| 917 |
+
},
|
| 918 |
+
{
|
| 919 |
+
"epoch": 0.3728325160507558,
|
| 920 |
+
"grad_norm": 4.907840728759766,
|
| 921 |
+
"learning_rate": 3.136749976302569e-05,
|
| 922 |
+
"loss": 4.3517,
|
| 923 |
+
"step": 11800
|
| 924 |
+
},
|
| 925 |
+
{
|
| 926 |
+
"epoch": 0.37599211364440627,
|
| 927 |
+
"grad_norm": 4.013961315155029,
|
| 928 |
+
"learning_rate": 3.12095168883693e-05,
|
| 929 |
+
"loss": 4.3636,
|
| 930 |
+
"step": 11900
|
| 931 |
+
},
|
| 932 |
+
{
|
| 933 |
+
"epoch": 0.37915171123805674,
|
| 934 |
+
"grad_norm": 3.6914377212524414,
|
| 935 |
+
"learning_rate": 3.105153401371291e-05,
|
| 936 |
+
"loss": 4.3087,
|
| 937 |
+
"step": 12000
|
| 938 |
+
},
|
| 939 |
+
{
|
| 940 |
+
"epoch": 0.37915171123805674,
|
| 941 |
+
"eval_loss": 4.48746395111084,
|
| 942 |
+
"eval_runtime": 715.6219,
|
| 943 |
+
"eval_samples_per_second": 122.853,
|
| 944 |
+
"eval_steps_per_second": 15.357,
|
| 945 |
+
"step": 12000
|
| 946 |
+
},
|
| 947 |
+
{
|
| 948 |
+
"epoch": 0.3823113088317072,
|
| 949 |
+
"grad_norm": 3.8867859840393066,
|
| 950 |
+
"learning_rate": 3.089355113905652e-05,
|
| 951 |
+
"loss": 4.337,
|
| 952 |
+
"step": 12100
|
| 953 |
+
},
|
| 954 |
+
{
|
| 955 |
+
"epoch": 0.38547090642535764,
|
| 956 |
+
"grad_norm": 4.648858070373535,
|
| 957 |
+
"learning_rate": 3.073556826440014e-05,
|
| 958 |
+
"loss": 4.3516,
|
| 959 |
+
"step": 12200
|
| 960 |
+
},
|
| 961 |
+
{
|
| 962 |
+
"epoch": 0.3886305040190081,
|
| 963 |
+
"grad_norm": 4.28641939163208,
|
| 964 |
+
"learning_rate": 3.0577585389743755e-05,
|
| 965 |
+
"loss": 4.3053,
|
| 966 |
+
"step": 12300
|
| 967 |
+
},
|
| 968 |
+
{
|
| 969 |
+
"epoch": 0.3917901016126586,
|
| 970 |
+
"grad_norm": 3.784064292907715,
|
| 971 |
+
"learning_rate": 3.0419602515087364e-05,
|
| 972 |
+
"loss": 4.3373,
|
| 973 |
+
"step": 12400
|
| 974 |
+
},
|
| 975 |
+
{
|
| 976 |
+
"epoch": 0.39494969920630907,
|
| 977 |
+
"grad_norm": 3.9344184398651123,
|
| 978 |
+
"learning_rate": 3.0261619640430976e-05,
|
| 979 |
+
"loss": 4.2861,
|
| 980 |
+
"step": 12500
|
| 981 |
+
},
|
| 982 |
+
{
|
| 983 |
+
"epoch": 0.39810929679995954,
|
| 984 |
+
"grad_norm": 4.234601974487305,
|
| 985 |
+
"learning_rate": 3.010363676577459e-05,
|
| 986 |
+
"loss": 4.3603,
|
| 987 |
+
"step": 12600
|
| 988 |
+
},
|
| 989 |
+
{
|
| 990 |
+
"epoch": 0.40126889439361,
|
| 991 |
+
"grad_norm": 4.8835835456848145,
|
| 992 |
+
"learning_rate": 2.9945653891118202e-05,
|
| 993 |
+
"loss": 4.3435,
|
| 994 |
+
"step": 12700
|
| 995 |
+
},
|
| 996 |
+
{
|
| 997 |
+
"epoch": 0.4044284919872605,
|
| 998 |
+
"grad_norm": 4.3561882972717285,
|
| 999 |
+
"learning_rate": 2.9787671016461814e-05,
|
| 1000 |
+
"loss": 4.3272,
|
| 1001 |
+
"step": 12800
|
| 1002 |
+
},
|
| 1003 |
+
{
|
| 1004 |
+
"epoch": 0.407588089580911,
|
| 1005 |
+
"grad_norm": 4.034718036651611,
|
| 1006 |
+
"learning_rate": 2.9629688141805427e-05,
|
| 1007 |
+
"loss": 4.2928,
|
| 1008 |
+
"step": 12900
|
| 1009 |
+
},
|
| 1010 |
+
{
|
| 1011 |
+
"epoch": 0.41074768717456145,
|
| 1012 |
+
"grad_norm": 4.587557315826416,
|
| 1013 |
+
"learning_rate": 2.947170526714904e-05,
|
| 1014 |
+
"loss": 4.3284,
|
| 1015 |
+
"step": 13000
|
| 1016 |
+
},
|
| 1017 |
+
{
|
| 1018 |
+
"epoch": 0.41074768717456145,
|
| 1019 |
+
"eval_loss": 4.477302074432373,
|
| 1020 |
+
"eval_runtime": 710.4818,
|
| 1021 |
+
"eval_samples_per_second": 123.741,
|
| 1022 |
+
"eval_steps_per_second": 15.468,
|
| 1023 |
+
"step": 13000
|
| 1024 |
+
},
|
| 1025 |
+
{
|
| 1026 |
+
"epoch": 0.4139072847682119,
|
| 1027 |
+
"grad_norm": 3.6614646911621094,
|
| 1028 |
+
"learning_rate": 2.9313722392492656e-05,
|
| 1029 |
+
"loss": 4.3301,
|
| 1030 |
+
"step": 13100
|
| 1031 |
+
},
|
| 1032 |
+
{
|
| 1033 |
+
"epoch": 0.4170668823618624,
|
| 1034 |
+
"grad_norm": 4.2207112312316895,
|
| 1035 |
+
"learning_rate": 2.915573951783627e-05,
|
| 1036 |
+
"loss": 4.3184,
|
| 1037 |
+
"step": 13200
|
| 1038 |
+
},
|
| 1039 |
+
{
|
| 1040 |
+
"epoch": 0.4202264799555129,
|
| 1041 |
+
"grad_norm": 4.571977138519287,
|
| 1042 |
+
"learning_rate": 2.8999336471926447e-05,
|
| 1043 |
+
"loss": 4.2802,
|
| 1044 |
+
"step": 13300
|
| 1045 |
+
},
|
| 1046 |
+
{
|
| 1047 |
+
"epoch": 0.42338607754916335,
|
| 1048 |
+
"grad_norm": 4.995270252227783,
|
| 1049 |
+
"learning_rate": 2.884135359727006e-05,
|
| 1050 |
+
"loss": 4.3286,
|
| 1051 |
+
"step": 13400
|
| 1052 |
+
},
|
| 1053 |
+
{
|
| 1054 |
+
"epoch": 0.42654567514281383,
|
| 1055 |
+
"grad_norm": 3.6203622817993164,
|
| 1056 |
+
"learning_rate": 2.8683370722613672e-05,
|
| 1057 |
+
"loss": 4.3566,
|
| 1058 |
+
"step": 13500
|
| 1059 |
+
},
|
| 1060 |
+
{
|
| 1061 |
+
"epoch": 0.4297052727364643,
|
| 1062 |
+
"grad_norm": 3.7246546745300293,
|
| 1063 |
+
"learning_rate": 2.852538784795728e-05,
|
| 1064 |
+
"loss": 4.3191,
|
| 1065 |
+
"step": 13600
|
| 1066 |
+
},
|
| 1067 |
+
{
|
| 1068 |
+
"epoch": 0.4328648703301148,
|
| 1069 |
+
"grad_norm": 4.118934631347656,
|
| 1070 |
+
"learning_rate": 2.8367404973300894e-05,
|
| 1071 |
+
"loss": 4.3338,
|
| 1072 |
+
"step": 13700
|
| 1073 |
+
},
|
| 1074 |
+
{
|
| 1075 |
+
"epoch": 0.4360244679237652,
|
| 1076 |
+
"grad_norm": 4.569579601287842,
|
| 1077 |
+
"learning_rate": 2.8209422098644506e-05,
|
| 1078 |
+
"loss": 4.3436,
|
| 1079 |
+
"step": 13800
|
| 1080 |
+
},
|
| 1081 |
+
{
|
| 1082 |
+
"epoch": 0.4391840655174157,
|
| 1083 |
+
"grad_norm": 4.280219554901123,
|
| 1084 |
+
"learning_rate": 2.805143922398812e-05,
|
| 1085 |
+
"loss": 4.2796,
|
| 1086 |
+
"step": 13900
|
| 1087 |
+
},
|
| 1088 |
+
{
|
| 1089 |
+
"epoch": 0.44234366311106615,
|
| 1090 |
+
"grad_norm": 4.866397857666016,
|
| 1091 |
+
"learning_rate": 2.7893456349331732e-05,
|
| 1092 |
+
"loss": 4.314,
|
| 1093 |
+
"step": 14000
|
| 1094 |
+
},
|
| 1095 |
+
{
|
| 1096 |
+
"epoch": 0.44234366311106615,
|
| 1097 |
+
"eval_loss": 4.468830585479736,
|
| 1098 |
+
"eval_runtime": 707.6161,
|
| 1099 |
+
"eval_samples_per_second": 124.243,
|
| 1100 |
+
"eval_steps_per_second": 15.531,
|
| 1101 |
+
"step": 14000
|
| 1102 |
+
},
|
| 1103 |
+
{
|
| 1104 |
+
"epoch": 0.44550326070471663,
|
| 1105 |
+
"grad_norm": 4.688275337219238,
|
| 1106 |
+
"learning_rate": 2.7735473474675344e-05,
|
| 1107 |
+
"loss": 4.2977,
|
| 1108 |
+
"step": 14100
|
| 1109 |
+
},
|
| 1110 |
+
{
|
| 1111 |
+
"epoch": 0.4486628582983671,
|
| 1112 |
+
"grad_norm": 4.715358734130859,
|
| 1113 |
+
"learning_rate": 2.7577490600018957e-05,
|
| 1114 |
+
"loss": 4.357,
|
| 1115 |
+
"step": 14200
|
| 1116 |
+
},
|
| 1117 |
+
{
|
| 1118 |
+
"epoch": 0.4518224558920176,
|
| 1119 |
+
"grad_norm": 4.193299293518066,
|
| 1120 |
+
"learning_rate": 2.741950772536257e-05,
|
| 1121 |
+
"loss": 4.3002,
|
| 1122 |
+
"step": 14300
|
| 1123 |
+
},
|
| 1124 |
+
{
|
| 1125 |
+
"epoch": 0.45498205348566806,
|
| 1126 |
+
"grad_norm": 5.159686088562012,
|
| 1127 |
+
"learning_rate": 2.7261524850706182e-05,
|
| 1128 |
+
"loss": 4.2935,
|
| 1129 |
+
"step": 14400
|
| 1130 |
+
},
|
| 1131 |
+
{
|
| 1132 |
+
"epoch": 0.45814165107931853,
|
| 1133 |
+
"grad_norm": 4.735175609588623,
|
| 1134 |
+
"learning_rate": 2.7103541976049795e-05,
|
| 1135 |
+
"loss": 4.3297,
|
| 1136 |
+
"step": 14500
|
| 1137 |
+
},
|
| 1138 |
+
{
|
| 1139 |
+
"epoch": 0.461301248672969,
|
| 1140 |
+
"grad_norm": 4.604960918426514,
|
| 1141 |
+
"learning_rate": 2.6945559101393408e-05,
|
| 1142 |
+
"loss": 4.308,
|
| 1143 |
+
"step": 14600
|
| 1144 |
+
},
|
| 1145 |
+
{
|
| 1146 |
+
"epoch": 0.4644608462666195,
|
| 1147 |
+
"grad_norm": 4.662102222442627,
|
| 1148 |
+
"learning_rate": 2.678757622673702e-05,
|
| 1149 |
+
"loss": 4.2911,
|
| 1150 |
+
"step": 14700
|
| 1151 |
+
},
|
| 1152 |
+
{
|
| 1153 |
+
"epoch": 0.46762044386026996,
|
| 1154 |
+
"grad_norm": 5.6145477294921875,
|
| 1155 |
+
"learning_rate": 2.6629593352080633e-05,
|
| 1156 |
+
"loss": 4.283,
|
| 1157 |
+
"step": 14800
|
| 1158 |
+
},
|
| 1159 |
+
{
|
| 1160 |
+
"epoch": 0.47078004145392044,
|
| 1161 |
+
"grad_norm": 5.1714606285095215,
|
| 1162 |
+
"learning_rate": 2.6471610477424246e-05,
|
| 1163 |
+
"loss": 4.3117,
|
| 1164 |
+
"step": 14900
|
| 1165 |
+
},
|
| 1166 |
+
{
|
| 1167 |
+
"epoch": 0.4739396390475709,
|
| 1168 |
+
"grad_norm": 4.684764385223389,
|
| 1169 |
+
"learning_rate": 2.6313627602767858e-05,
|
| 1170 |
+
"loss": 4.3159,
|
| 1171 |
+
"step": 15000
|
| 1172 |
+
},
|
| 1173 |
+
{
|
| 1174 |
+
"epoch": 0.4739396390475709,
|
| 1175 |
+
"eval_loss": 4.460094451904297,
|
| 1176 |
+
"eval_runtime": 708.6547,
|
| 1177 |
+
"eval_samples_per_second": 124.06,
|
| 1178 |
+
"eval_steps_per_second": 15.508,
|
| 1179 |
+
"step": 15000
|
| 1180 |
+
},
|
| 1181 |
+
{
|
| 1182 |
+
"epoch": 0.4770992366412214,
|
| 1183 |
+
"grad_norm": 4.812125205993652,
|
| 1184 |
+
"learning_rate": 2.6155644728111474e-05,
|
| 1185 |
+
"loss": 4.297,
|
| 1186 |
+
"step": 15100
|
| 1187 |
+
},
|
| 1188 |
+
{
|
| 1189 |
+
"epoch": 0.48025883423487187,
|
| 1190 |
+
"grad_norm": 4.402428150177002,
|
| 1191 |
+
"learning_rate": 2.5997661853455087e-05,
|
| 1192 |
+
"loss": 4.2707,
|
| 1193 |
+
"step": 15200
|
| 1194 |
+
},
|
| 1195 |
+
{
|
| 1196 |
+
"epoch": 0.48341843182852234,
|
| 1197 |
+
"grad_norm": 3.6931350231170654,
|
| 1198 |
+
"learning_rate": 2.5841258807545265e-05,
|
| 1199 |
+
"loss": 4.2843,
|
| 1200 |
+
"step": 15300
|
| 1201 |
+
},
|
| 1202 |
+
{
|
| 1203 |
+
"epoch": 0.4865780294221728,
|
| 1204 |
+
"grad_norm": 4.092090129852295,
|
| 1205 |
+
"learning_rate": 2.5683275932888878e-05,
|
| 1206 |
+
"loss": 4.2672,
|
| 1207 |
+
"step": 15400
|
| 1208 |
+
},
|
| 1209 |
+
{
|
| 1210 |
+
"epoch": 0.48973762701582324,
|
| 1211 |
+
"grad_norm": 4.624776363372803,
|
| 1212 |
+
"learning_rate": 2.552529305823249e-05,
|
| 1213 |
+
"loss": 4.3225,
|
| 1214 |
+
"step": 15500
|
| 1215 |
+
},
|
| 1216 |
+
{
|
| 1217 |
+
"epoch": 0.4928972246094737,
|
| 1218 |
+
"grad_norm": 4.362007141113281,
|
| 1219 |
+
"learning_rate": 2.5367310183576103e-05,
|
| 1220 |
+
"loss": 4.3164,
|
| 1221 |
+
"step": 15600
|
| 1222 |
+
},
|
| 1223 |
+
{
|
| 1224 |
+
"epoch": 0.4960568222031242,
|
| 1225 |
+
"grad_norm": 3.67773175239563,
|
| 1226 |
+
"learning_rate": 2.5209327308919716e-05,
|
| 1227 |
+
"loss": 4.2915,
|
| 1228 |
+
"step": 15700
|
| 1229 |
+
},
|
| 1230 |
+
{
|
| 1231 |
+
"epoch": 0.49921641979677467,
|
| 1232 |
+
"grad_norm": 3.7847766876220703,
|
| 1233 |
+
"learning_rate": 2.505134443426333e-05,
|
| 1234 |
+
"loss": 4.2696,
|
| 1235 |
+
"step": 15800
|
| 1236 |
+
},
|
| 1237 |
+
{
|
| 1238 |
+
"epoch": 0.5023760173904251,
|
| 1239 |
+
"grad_norm": 4.13474178314209,
|
| 1240 |
+
"learning_rate": 2.489336155960694e-05,
|
| 1241 |
+
"loss": 4.3031,
|
| 1242 |
+
"step": 15900
|
| 1243 |
+
},
|
| 1244 |
+
{
|
| 1245 |
+
"epoch": 0.5055356149840756,
|
| 1246 |
+
"grad_norm": 4.143453598022461,
|
| 1247 |
+
"learning_rate": 2.4735378684950554e-05,
|
| 1248 |
+
"loss": 4.3384,
|
| 1249 |
+
"step": 16000
|
| 1250 |
+
},
|
| 1251 |
+
{
|
| 1252 |
+
"epoch": 0.5055356149840756,
|
| 1253 |
+
"eval_loss": 4.452681541442871,
|
| 1254 |
+
"eval_runtime": 706.2531,
|
| 1255 |
+
"eval_samples_per_second": 124.482,
|
| 1256 |
+
"eval_steps_per_second": 15.561,
|
| 1257 |
+
"step": 16000
|
| 1258 |
+
},
|
| 1259 |
+
{
|
| 1260 |
+
"epoch": 0.5086952125777261,
|
| 1261 |
+
"grad_norm": 4.447348594665527,
|
| 1262 |
+
"learning_rate": 2.4577395810294166e-05,
|
| 1263 |
+
"loss": 4.2958,
|
| 1264 |
+
"step": 16100
|
| 1265 |
+
},
|
| 1266 |
+
{
|
| 1267 |
+
"epoch": 0.5118548101713766,
|
| 1268 |
+
"grad_norm": 4.753751754760742,
|
| 1269 |
+
"learning_rate": 2.441941293563778e-05,
|
| 1270 |
+
"loss": 4.3129,
|
| 1271 |
+
"step": 16200
|
| 1272 |
+
},
|
| 1273 |
+
{
|
| 1274 |
+
"epoch": 0.515014407765027,
|
| 1275 |
+
"grad_norm": 3.83662748336792,
|
| 1276 |
+
"learning_rate": 2.426143006098139e-05,
|
| 1277 |
+
"loss": 4.3168,
|
| 1278 |
+
"step": 16300
|
| 1279 |
+
},
|
| 1280 |
+
{
|
| 1281 |
+
"epoch": 0.5181740053586775,
|
| 1282 |
+
"grad_norm": 3.786252737045288,
|
| 1283 |
+
"learning_rate": 2.4103447186325004e-05,
|
| 1284 |
+
"loss": 4.2844,
|
| 1285 |
+
"step": 16400
|
| 1286 |
+
},
|
| 1287 |
+
{
|
| 1288 |
+
"epoch": 0.521333602952328,
|
| 1289 |
+
"grad_norm": 3.9351420402526855,
|
| 1290 |
+
"learning_rate": 2.3945464311668617e-05,
|
| 1291 |
+
"loss": 4.2816,
|
| 1292 |
+
"step": 16500
|
| 1293 |
+
},
|
| 1294 |
+
{
|
| 1295 |
+
"epoch": 0.5244932005459785,
|
| 1296 |
+
"grad_norm": 4.529867172241211,
|
| 1297 |
+
"learning_rate": 2.378748143701223e-05,
|
| 1298 |
+
"loss": 4.302,
|
| 1299 |
+
"step": 16600
|
| 1300 |
+
},
|
| 1301 |
+
{
|
| 1302 |
+
"epoch": 0.527652798139629,
|
| 1303 |
+
"grad_norm": 4.681921005249023,
|
| 1304 |
+
"learning_rate": 2.3629498562355842e-05,
|
| 1305 |
+
"loss": 4.2904,
|
| 1306 |
+
"step": 16700
|
| 1307 |
+
},
|
| 1308 |
+
{
|
| 1309 |
+
"epoch": 0.5308123957332794,
|
| 1310 |
+
"grad_norm": 3.3952362537384033,
|
| 1311 |
+
"learning_rate": 2.3471515687699455e-05,
|
| 1312 |
+
"loss": 4.2876,
|
| 1313 |
+
"step": 16800
|
| 1314 |
+
},
|
| 1315 |
+
{
|
| 1316 |
+
"epoch": 0.5339719933269299,
|
| 1317 |
+
"grad_norm": 4.524105548858643,
|
| 1318 |
+
"learning_rate": 2.3313532813043067e-05,
|
| 1319 |
+
"loss": 4.262,
|
| 1320 |
+
"step": 16900
|
| 1321 |
+
},
|
| 1322 |
+
{
|
| 1323 |
+
"epoch": 0.5371315909205804,
|
| 1324 |
+
"grad_norm": 4.248294830322266,
|
| 1325 |
+
"learning_rate": 2.315554993838668e-05,
|
| 1326 |
+
"loss": 4.2975,
|
| 1327 |
+
"step": 17000
|
| 1328 |
+
},
|
| 1329 |
+
{
|
| 1330 |
+
"epoch": 0.5371315909205804,
|
| 1331 |
+
"eval_loss": 4.445527076721191,
|
| 1332 |
+
"eval_runtime": 705.3764,
|
| 1333 |
+
"eval_samples_per_second": 124.637,
|
| 1334 |
+
"eval_steps_per_second": 15.58,
|
| 1335 |
+
"step": 17000
|
| 1336 |
+
},
|
| 1337 |
+
{
|
| 1338 |
+
"epoch": 0.5402911885142309,
|
| 1339 |
+
"grad_norm": 4.230813026428223,
|
| 1340 |
+
"learning_rate": 2.2997567063730293e-05,
|
| 1341 |
+
"loss": 4.2806,
|
| 1342 |
+
"step": 17100
|
| 1343 |
+
},
|
| 1344 |
+
{
|
| 1345 |
+
"epoch": 0.5434507861078813,
|
| 1346 |
+
"grad_norm": 3.7945194244384766,
|
| 1347 |
+
"learning_rate": 2.2839584189073905e-05,
|
| 1348 |
+
"loss": 4.2942,
|
| 1349 |
+
"step": 17200
|
| 1350 |
+
},
|
| 1351 |
+
{
|
| 1352 |
+
"epoch": 0.5466103837015318,
|
| 1353 |
+
"grad_norm": 4.855790138244629,
|
| 1354 |
+
"learning_rate": 2.2681601314417518e-05,
|
| 1355 |
+
"loss": 4.2818,
|
| 1356 |
+
"step": 17300
|
| 1357 |
+
},
|
| 1358 |
+
{
|
| 1359 |
+
"epoch": 0.5497699812951823,
|
| 1360 |
+
"grad_norm": 4.7962236404418945,
|
| 1361 |
+
"learning_rate": 2.252361843976113e-05,
|
| 1362 |
+
"loss": 4.2804,
|
| 1363 |
+
"step": 17400
|
| 1364 |
+
},
|
| 1365 |
+
{
|
| 1366 |
+
"epoch": 0.5529295788888328,
|
| 1367 |
+
"grad_norm": 4.337451934814453,
|
| 1368 |
+
"learning_rate": 2.2365635565104747e-05,
|
| 1369 |
+
"loss": 4.3063,
|
| 1370 |
+
"step": 17500
|
| 1371 |
+
},
|
| 1372 |
+
{
|
| 1373 |
+
"epoch": 0.5560891764824832,
|
| 1374 |
+
"grad_norm": 4.422916889190674,
|
| 1375 |
+
"learning_rate": 2.220765269044836e-05,
|
| 1376 |
+
"loss": 4.2942,
|
| 1377 |
+
"step": 17600
|
| 1378 |
+
},
|
| 1379 |
+
{
|
| 1380 |
+
"epoch": 0.5592487740761337,
|
| 1381 |
+
"grad_norm": 4.27534294128418,
|
| 1382 |
+
"learning_rate": 2.204966981579197e-05,
|
| 1383 |
+
"loss": 4.2828,
|
| 1384 |
+
"step": 17700
|
| 1385 |
+
},
|
| 1386 |
+
{
|
| 1387 |
+
"epoch": 0.5624083716697842,
|
| 1388 |
+
"grad_norm": 4.3357648849487305,
|
| 1389 |
+
"learning_rate": 2.1893266769882147e-05,
|
| 1390 |
+
"loss": 4.2766,
|
| 1391 |
+
"step": 17800
|
| 1392 |
+
},
|
| 1393 |
+
{
|
| 1394 |
+
"epoch": 0.5655679692634346,
|
| 1395 |
+
"grad_norm": 3.859369993209839,
|
| 1396 |
+
"learning_rate": 2.173528389522576e-05,
|
| 1397 |
+
"loss": 4.2979,
|
| 1398 |
+
"step": 17900
|
| 1399 |
+
},
|
| 1400 |
+
{
|
| 1401 |
+
"epoch": 0.568727566857085,
|
| 1402 |
+
"grad_norm": 5.381210803985596,
|
| 1403 |
+
"learning_rate": 2.1577301020569372e-05,
|
| 1404 |
+
"loss": 4.3169,
|
| 1405 |
+
"step": 18000
|
| 1406 |
+
},
|
| 1407 |
+
{
|
| 1408 |
+
"epoch": 0.568727566857085,
|
| 1409 |
+
"eval_loss": 4.439344882965088,
|
| 1410 |
+
"eval_runtime": 708.2829,
|
| 1411 |
+
"eval_samples_per_second": 124.126,
|
| 1412 |
+
"eval_steps_per_second": 15.516,
|
| 1413 |
+
"step": 18000
|
| 1414 |
+
},
|
| 1415 |
+
{
|
| 1416 |
+
"epoch": 0.5718871644507355,
|
| 1417 |
+
"grad_norm": 3.3209519386291504,
|
| 1418 |
+
"learning_rate": 2.1419318145912985e-05,
|
| 1419 |
+
"loss": 4.2588,
|
| 1420 |
+
"step": 18100
|
| 1421 |
+
},
|
| 1422 |
+
{
|
| 1423 |
+
"epoch": 0.575046762044386,
|
| 1424 |
+
"grad_norm": 3.9373040199279785,
|
| 1425 |
+
"learning_rate": 2.1261335271256597e-05,
|
| 1426 |
+
"loss": 4.2895,
|
| 1427 |
+
"step": 18200
|
| 1428 |
+
},
|
| 1429 |
+
{
|
| 1430 |
+
"epoch": 0.5782063596380365,
|
| 1431 |
+
"grad_norm": 3.453148126602173,
|
| 1432 |
+
"learning_rate": 2.110335239660021e-05,
|
| 1433 |
+
"loss": 4.276,
|
| 1434 |
+
"step": 18300
|
| 1435 |
+
},
|
| 1436 |
+
{
|
| 1437 |
+
"epoch": 0.5813659572316869,
|
| 1438 |
+
"grad_norm": 3.7350664138793945,
|
| 1439 |
+
"learning_rate": 2.0945369521943823e-05,
|
| 1440 |
+
"loss": 4.3268,
|
| 1441 |
+
"step": 18400
|
| 1442 |
+
},
|
| 1443 |
+
{
|
| 1444 |
+
"epoch": 0.5845255548253374,
|
| 1445 |
+
"grad_norm": 4.277946949005127,
|
| 1446 |
+
"learning_rate": 2.0787386647287435e-05,
|
| 1447 |
+
"loss": 4.2573,
|
| 1448 |
+
"step": 18500
|
| 1449 |
+
},
|
| 1450 |
+
{
|
| 1451 |
+
"epoch": 0.5876851524189879,
|
| 1452 |
+
"grad_norm": 4.000477313995361,
|
| 1453 |
+
"learning_rate": 2.0629403772631048e-05,
|
| 1454 |
+
"loss": 4.2568,
|
| 1455 |
+
"step": 18600
|
| 1456 |
+
},
|
| 1457 |
+
{
|
| 1458 |
+
"epoch": 0.5908447500126384,
|
| 1459 |
+
"grad_norm": 3.6891930103302,
|
| 1460 |
+
"learning_rate": 2.047142089797466e-05,
|
| 1461 |
+
"loss": 4.2784,
|
| 1462 |
+
"step": 18700
|
| 1463 |
+
},
|
| 1464 |
+
{
|
| 1465 |
+
"epoch": 0.5940043476062888,
|
| 1466 |
+
"grad_norm": 4.800564765930176,
|
| 1467 |
+
"learning_rate": 2.0313438023318273e-05,
|
| 1468 |
+
"loss": 4.3016,
|
| 1469 |
+
"step": 18800
|
| 1470 |
+
},
|
| 1471 |
+
{
|
| 1472 |
+
"epoch": 0.5971639451999393,
|
| 1473 |
+
"grad_norm": 5.003584384918213,
|
| 1474 |
+
"learning_rate": 2.0155455148661886e-05,
|
| 1475 |
+
"loss": 4.2818,
|
| 1476 |
+
"step": 18900
|
| 1477 |
+
},
|
| 1478 |
+
{
|
| 1479 |
+
"epoch": 0.6003235427935898,
|
| 1480 |
+
"grad_norm": 4.17577600479126,
|
| 1481 |
+
"learning_rate": 1.99974722740055e-05,
|
| 1482 |
+
"loss": 4.2794,
|
| 1483 |
+
"step": 19000
|
| 1484 |
+
},
|
| 1485 |
+
{
|
| 1486 |
+
"epoch": 0.6003235427935898,
|
| 1487 |
+
"eval_loss": 4.433346748352051,
|
| 1488 |
+
"eval_runtime": 712.1597,
|
| 1489 |
+
"eval_samples_per_second": 123.45,
|
| 1490 |
+
"eval_steps_per_second": 15.432,
|
| 1491 |
+
"step": 19000
|
| 1492 |
+
},
|
| 1493 |
+
{
|
| 1494 |
+
"epoch": 0.6034831403872403,
|
| 1495 |
+
"grad_norm": 4.12880277633667,
|
| 1496 |
+
"learning_rate": 1.983948939934911e-05,
|
| 1497 |
+
"loss": 4.2539,
|
| 1498 |
+
"step": 19100
|
| 1499 |
+
},
|
| 1500 |
+
{
|
| 1501 |
+
"epoch": 0.6066427379808907,
|
| 1502 |
+
"grad_norm": 4.599102020263672,
|
| 1503 |
+
"learning_rate": 1.9681506524692724e-05,
|
| 1504 |
+
"loss": 4.2842,
|
| 1505 |
+
"step": 19200
|
| 1506 |
+
},
|
| 1507 |
+
{
|
| 1508 |
+
"epoch": 0.6098023355745412,
|
| 1509 |
+
"grad_norm": 4.2362380027771,
|
| 1510 |
+
"learning_rate": 1.9523523650036336e-05,
|
| 1511 |
+
"loss": 4.2803,
|
| 1512 |
+
"step": 19300
|
| 1513 |
+
},
|
| 1514 |
+
{
|
| 1515 |
+
"epoch": 0.6129619331681917,
|
| 1516 |
+
"grad_norm": 4.119621753692627,
|
| 1517 |
+
"learning_rate": 1.936554077537995e-05,
|
| 1518 |
+
"loss": 4.2413,
|
| 1519 |
+
"step": 19400
|
| 1520 |
+
},
|
| 1521 |
+
{
|
| 1522 |
+
"epoch": 0.6161215307618422,
|
| 1523 |
+
"grad_norm": 4.374582290649414,
|
| 1524 |
+
"learning_rate": 1.9207557900723565e-05,
|
| 1525 |
+
"loss": 4.3056,
|
| 1526 |
+
"step": 19500
|
| 1527 |
+
},
|
| 1528 |
+
{
|
| 1529 |
+
"epoch": 0.6192811283554926,
|
| 1530 |
+
"grad_norm": 3.5877952575683594,
|
| 1531 |
+
"learning_rate": 1.9049575026067178e-05,
|
| 1532 |
+
"loss": 4.2978,
|
| 1533 |
+
"step": 19600
|
| 1534 |
+
},
|
| 1535 |
+
{
|
| 1536 |
+
"epoch": 0.6224407259491431,
|
| 1537 |
+
"grad_norm": 4.062015056610107,
|
| 1538 |
+
"learning_rate": 1.889159215141079e-05,
|
| 1539 |
+
"loss": 4.2525,
|
| 1540 |
+
"step": 19700
|
| 1541 |
+
},
|
| 1542 |
+
{
|
| 1543 |
+
"epoch": 0.6256003235427936,
|
| 1544 |
+
"grad_norm": 4.872693061828613,
|
| 1545 |
+
"learning_rate": 1.8733609276754403e-05,
|
| 1546 |
+
"loss": 4.3039,
|
| 1547 |
+
"step": 19800
|
| 1548 |
+
},
|
| 1549 |
+
{
|
| 1550 |
+
"epoch": 0.6287599211364441,
|
| 1551 |
+
"grad_norm": 3.899491310119629,
|
| 1552 |
+
"learning_rate": 1.8575626402098016e-05,
|
| 1553 |
+
"loss": 4.3016,
|
| 1554 |
+
"step": 19900
|
| 1555 |
+
},
|
| 1556 |
+
{
|
| 1557 |
+
"epoch": 0.6319195187300946,
|
| 1558 |
+
"grad_norm": 3.927971601486206,
|
| 1559 |
+
"learning_rate": 1.8417643527441628e-05,
|
| 1560 |
+
"loss": 4.2883,
|
| 1561 |
+
"step": 20000
|
| 1562 |
+
},
|
| 1563 |
+
{
|
| 1564 |
+
"epoch": 0.6319195187300946,
|
| 1565 |
+
"eval_loss": 4.4285101890563965,
|
| 1566 |
+
"eval_runtime": 708.5523,
|
| 1567 |
+
"eval_samples_per_second": 124.078,
|
| 1568 |
+
"eval_steps_per_second": 15.51,
|
| 1569 |
+
"step": 20000
|
| 1570 |
+
}
|
| 1571 |
+
],
|
| 1572 |
+
"logging_steps": 100,
|
| 1573 |
+
"max_steps": 31649,
|
| 1574 |
+
"num_input_tokens_seen": 0,
|
| 1575 |
+
"num_train_epochs": 1,
|
| 1576 |
+
"save_steps": 10000,
|
| 1577 |
+
"stateful_callbacks": {
|
| 1578 |
+
"TrainerControl": {
|
| 1579 |
+
"args": {
|
| 1580 |
+
"should_epoch_stop": false,
|
| 1581 |
+
"should_evaluate": false,
|
| 1582 |
+
"should_log": false,
|
| 1583 |
+
"should_save": true,
|
| 1584 |
+
"should_training_stop": false
|
| 1585 |
+
},
|
| 1586 |
+
"attributes": {}
|
| 1587 |
+
}
|
| 1588 |
+
},
|
| 1589 |
+
"total_flos": 9.1623063552e+16,
|
| 1590 |
+
"train_batch_size": 5,
|
| 1591 |
+
"trial_name": null,
|
| 1592 |
+
"trial_params": null
|
| 1593 |
+
}
|
fine_tuned_model_output_1/checkpoint-20000/training_args.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:95acebf5e801652183ac4ad483ff8bfe2ede739795ede7240bbee3b5a7707481
|
| 3 |
+
size 5368
|
fine_tuned_model_output_1/checkpoint-30000/config.json
ADDED
|
@@ -0,0 +1,30 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"_name_or_path": "./linked_model_with_custom_tokenizer",
|
| 3 |
+
"architectures": [
|
| 4 |
+
"Qwen2ForCausalLM"
|
| 5 |
+
],
|
| 6 |
+
"attention_dropout": 0.0,
|
| 7 |
+
"bos_token_id": 151643,
|
| 8 |
+
"eos_token_id": 151643,
|
| 9 |
+
"hidden_act": "silu",
|
| 10 |
+
"hidden_size": 896,
|
| 11 |
+
"initializer_range": 0.02,
|
| 12 |
+
"intermediate_size": 4864,
|
| 13 |
+
"max_position_embeddings": 32768,
|
| 14 |
+
"max_window_layers": 24,
|
| 15 |
+
"model_type": "qwen2",
|
| 16 |
+
"num_attention_heads": 14,
|
| 17 |
+
"num_hidden_layers": 4,
|
| 18 |
+
"num_key_value_heads": 2,
|
| 19 |
+
"rms_norm_eps": 1e-06,
|
| 20 |
+
"rope_scaling": null,
|
| 21 |
+
"rope_theta": 1000000.0,
|
| 22 |
+
"sliding_window": 32768,
|
| 23 |
+
"tie_word_embeddings": true,
|
| 24 |
+
"torch_dtype": "float32",
|
| 25 |
+
"transformers_version": "4.49.0",
|
| 26 |
+
"use_cache": true,
|
| 27 |
+
"use_mrope": false,
|
| 28 |
+
"use_sliding_window": false,
|
| 29 |
+
"vocab_size": 32002
|
| 30 |
+
}
|
fine_tuned_model_output_1/checkpoint-30000/generation_config.json
ADDED
|
@@ -0,0 +1,6 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"_from_model_config": true,
|
| 3 |
+
"bos_token_id": 151643,
|
| 4 |
+
"eos_token_id": 151643,
|
| 5 |
+
"transformers_version": "4.49.0"
|
| 6 |
+
}
|
fine_tuned_model_output_1/checkpoint-30000/model.safetensors
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:6eecfa0d0e62ec5cd8ced14fd4c791e6497c40e9430a1fd400c788ff67807064
|
| 3 |
+
size 353302384
|
fine_tuned_model_output_1/checkpoint-30000/optimizer.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:2da3c697ccd03203f23a35da06b9bc1f663286eea527fa0f324f5a918b188ffe
|
| 3 |
+
size 706636090
|
fine_tuned_model_output_1/checkpoint-30000/rng_state.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:00b1a4b7d78ed052f3955ac3d0a9db8fc337dc97611574c597610453d8d99939
|
| 3 |
+
size 14244
|
fine_tuned_model_output_1/checkpoint-30000/scaler.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:a61ca95a871157454347a7bd2c706cc3caee4e5a038f59ddd7be0ba6b00778ba
|
| 3 |
+
size 988
|
fine_tuned_model_output_1/checkpoint-30000/scheduler.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:935c57b46c3adb70ba3f7c4e3026b6367b01986fe70fdc1bddbd074c443d1170
|
| 3 |
+
size 1064
|
fine_tuned_model_output_1/checkpoint-30000/special_tokens_map.json
ADDED
|
@@ -0,0 +1,30 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"bos_token": {
|
| 3 |
+
"content": "<s>",
|
| 4 |
+
"lstrip": false,
|
| 5 |
+
"normalized": false,
|
| 6 |
+
"rstrip": false,
|
| 7 |
+
"single_word": false
|
| 8 |
+
},
|
| 9 |
+
"eos_token": {
|
| 10 |
+
"content": "</s>",
|
| 11 |
+
"lstrip": false,
|
| 12 |
+
"normalized": false,
|
| 13 |
+
"rstrip": false,
|
| 14 |
+
"single_word": false
|
| 15 |
+
},
|
| 16 |
+
"pad_token": {
|
| 17 |
+
"content": "</s>",
|
| 18 |
+
"lstrip": false,
|
| 19 |
+
"normalized": false,
|
| 20 |
+
"rstrip": false,
|
| 21 |
+
"single_word": false
|
| 22 |
+
},
|
| 23 |
+
"unk_token": {
|
| 24 |
+
"content": "<unk>",
|
| 25 |
+
"lstrip": false,
|
| 26 |
+
"normalized": false,
|
| 27 |
+
"rstrip": false,
|
| 28 |
+
"single_word": false
|
| 29 |
+
}
|
| 30 |
+
}
|
fine_tuned_model_output_1/checkpoint-30000/tokenizer.json
ADDED
|
The diff for this file is too large to render.
See raw diff
|
|
|
fine_tuned_model_output_1/checkpoint-30000/tokenizer_config.json
ADDED
|
@@ -0,0 +1,67 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"added_tokens_decoder": {
|
| 3 |
+
"0": {
|
| 4 |
+
"content": "<unk>",
|
| 5 |
+
"lstrip": false,
|
| 6 |
+
"normalized": false,
|
| 7 |
+
"rstrip": false,
|
| 8 |
+
"single_word": false,
|
| 9 |
+
"special": true
|
| 10 |
+
},
|
| 11 |
+
"1": {
|
| 12 |
+
"content": "<s>",
|
| 13 |
+
"lstrip": false,
|
| 14 |
+
"normalized": false,
|
| 15 |
+
"rstrip": false,
|
| 16 |
+
"single_word": false,
|
| 17 |
+
"special": true
|
| 18 |
+
},
|
| 19 |
+
"2": {
|
| 20 |
+
"content": "</s>",
|
| 21 |
+
"lstrip": false,
|
| 22 |
+
"normalized": false,
|
| 23 |
+
"rstrip": false,
|
| 24 |
+
"single_word": false,
|
| 25 |
+
"special": true
|
| 26 |
+
},
|
| 27 |
+
"3": {
|
| 28 |
+
"content": "<pad>",
|
| 29 |
+
"lstrip": false,
|
| 30 |
+
"normalized": false,
|
| 31 |
+
"rstrip": false,
|
| 32 |
+
"single_word": false,
|
| 33 |
+
"special": true
|
| 34 |
+
},
|
| 35 |
+
"32000": {
|
| 36 |
+
"content": "Sanskrit:",
|
| 37 |
+
"lstrip": false,
|
| 38 |
+
"normalized": true,
|
| 39 |
+
"rstrip": false,
|
| 40 |
+
"single_word": false,
|
| 41 |
+
"special": false
|
| 42 |
+
},
|
| 43 |
+
"32001": {
|
| 44 |
+
"content": "English:",
|
| 45 |
+
"lstrip": false,
|
| 46 |
+
"normalized": true,
|
| 47 |
+
"rstrip": false,
|
| 48 |
+
"single_word": false,
|
| 49 |
+
"special": false
|
| 50 |
+
}
|
| 51 |
+
},
|
| 52 |
+
"bos_token": "<s>",
|
| 53 |
+
"clean_up_tokenization_spaces": false,
|
| 54 |
+
"eos_token": "</s>",
|
| 55 |
+
"extra_special_tokens": {},
|
| 56 |
+
"max_length": 512,
|
| 57 |
+
"model_max_length": 1000000000000000019884624838656,
|
| 58 |
+
"pad_to_multiple_of": null,
|
| 59 |
+
"pad_token": "</s>",
|
| 60 |
+
"pad_token_type_id": 0,
|
| 61 |
+
"padding_side": "right",
|
| 62 |
+
"stride": 0,
|
| 63 |
+
"tokenizer_class": "PreTrainedTokenizer",
|
| 64 |
+
"truncation_side": "right",
|
| 65 |
+
"truncation_strategy": "longest_first",
|
| 66 |
+
"unk_token": "<unk>"
|
| 67 |
+
}
|
fine_tuned_model_output_1/checkpoint-30000/trainer_state.json
ADDED
|
@@ -0,0 +1,2373 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": 4.397016525268555,
|
| 3 |
+
"best_model_checkpoint": "./fine_tuned_model_output_1\\checkpoint-30000",
|
| 4 |
+
"epoch": 0.9478792780951418,
|
| 5 |
+
"eval_steps": 1000,
|
| 6 |
+
"global_step": 30000,
|
| 7 |
+
"is_hyper_param_search": false,
|
| 8 |
+
"is_local_process_zero": true,
|
| 9 |
+
"is_world_process_zero": true,
|
| 10 |
+
"log_history": [
|
| 11 |
+
{
|
| 12 |
+
"epoch": 0.0031595975936504726,
|
| 13 |
+
"grad_norm": 5.00691556930542,
|
| 14 |
+
"learning_rate": 4.984517678283674e-05,
|
| 15 |
+
"loss": 5.1637,
|
| 16 |
+
"step": 100
|
| 17 |
+
},
|
| 18 |
+
{
|
| 19 |
+
"epoch": 0.006319195187300945,
|
| 20 |
+
"grad_norm": 4.790140628814697,
|
| 21 |
+
"learning_rate": 4.968719390818036e-05,
|
| 22 |
+
"loss": 4.9551,
|
| 23 |
+
"step": 200
|
| 24 |
+
},
|
| 25 |
+
{
|
| 26 |
+
"epoch": 0.009478792780951418,
|
| 27 |
+
"grad_norm": 4.75383186340332,
|
| 28 |
+
"learning_rate": 4.952921103352397e-05,
|
| 29 |
+
"loss": 4.8786,
|
| 30 |
+
"step": 300
|
| 31 |
+
},
|
| 32 |
+
{
|
| 33 |
+
"epoch": 0.01263839037460189,
|
| 34 |
+
"grad_norm": 5.218325138092041,
|
| 35 |
+
"learning_rate": 4.9371228158867584e-05,
|
| 36 |
+
"loss": 4.8117,
|
| 37 |
+
"step": 400
|
| 38 |
+
},
|
| 39 |
+
{
|
| 40 |
+
"epoch": 0.015797987968252363,
|
| 41 |
+
"grad_norm": 5.101843357086182,
|
| 42 |
+
"learning_rate": 4.9213245284211194e-05,
|
| 43 |
+
"loss": 4.8153,
|
| 44 |
+
"step": 500
|
| 45 |
+
},
|
| 46 |
+
{
|
| 47 |
+
"epoch": 0.018957585561902836,
|
| 48 |
+
"grad_norm": 4.165743350982666,
|
| 49 |
+
"learning_rate": 4.905526240955481e-05,
|
| 50 |
+
"loss": 4.7331,
|
| 51 |
+
"step": 600
|
| 52 |
+
},
|
| 53 |
+
{
|
| 54 |
+
"epoch": 0.02211718315555331,
|
| 55 |
+
"grad_norm": 4.386114597320557,
|
| 56 |
+
"learning_rate": 4.889727953489842e-05,
|
| 57 |
+
"loss": 4.7982,
|
| 58 |
+
"step": 700
|
| 59 |
+
},
|
| 60 |
+
{
|
| 61 |
+
"epoch": 0.02527678074920378,
|
| 62 |
+
"grad_norm": 4.677374362945557,
|
| 63 |
+
"learning_rate": 4.8739296660242035e-05,
|
| 64 |
+
"loss": 4.751,
|
| 65 |
+
"step": 800
|
| 66 |
+
},
|
| 67 |
+
{
|
| 68 |
+
"epoch": 0.028436378342854254,
|
| 69 |
+
"grad_norm": 4.1365814208984375,
|
| 70 |
+
"learning_rate": 4.8581313785585644e-05,
|
| 71 |
+
"loss": 4.7631,
|
| 72 |
+
"step": 900
|
| 73 |
+
},
|
| 74 |
+
{
|
| 75 |
+
"epoch": 0.031595975936504726,
|
| 76 |
+
"grad_norm": 4.651081562042236,
|
| 77 |
+
"learning_rate": 4.842333091092926e-05,
|
| 78 |
+
"loss": 4.7504,
|
| 79 |
+
"step": 1000
|
| 80 |
+
},
|
| 81 |
+
{
|
| 82 |
+
"epoch": 0.031595975936504726,
|
| 83 |
+
"eval_loss": 4.873119354248047,
|
| 84 |
+
"eval_runtime": 665.3679,
|
| 85 |
+
"eval_samples_per_second": 132.131,
|
| 86 |
+
"eval_steps_per_second": 16.517,
|
| 87 |
+
"step": 1000
|
| 88 |
+
},
|
| 89 |
+
{
|
| 90 |
+
"epoch": 0.0347555735301552,
|
| 91 |
+
"grad_norm": 5.823278903961182,
|
| 92 |
+
"learning_rate": 4.826534803627287e-05,
|
| 93 |
+
"loss": 4.6999,
|
| 94 |
+
"step": 1100
|
| 95 |
+
},
|
| 96 |
+
{
|
| 97 |
+
"epoch": 0.03791517112380567,
|
| 98 |
+
"grad_norm": 4.6261773109436035,
|
| 99 |
+
"learning_rate": 4.8107365161616486e-05,
|
| 100 |
+
"loss": 4.6704,
|
| 101 |
+
"step": 1200
|
| 102 |
+
},
|
| 103 |
+
{
|
| 104 |
+
"epoch": 0.04107476871745615,
|
| 105 |
+
"grad_norm": 4.6412200927734375,
|
| 106 |
+
"learning_rate": 4.79493822869601e-05,
|
| 107 |
+
"loss": 4.691,
|
| 108 |
+
"step": 1300
|
| 109 |
+
},
|
| 110 |
+
{
|
| 111 |
+
"epoch": 0.04423436631110662,
|
| 112 |
+
"grad_norm": 4.368412494659424,
|
| 113 |
+
"learning_rate": 4.779139941230371e-05,
|
| 114 |
+
"loss": 4.7048,
|
| 115 |
+
"step": 1400
|
| 116 |
+
},
|
| 117 |
+
{
|
| 118 |
+
"epoch": 0.04739396390475709,
|
| 119 |
+
"grad_norm": 3.9410157203674316,
|
| 120 |
+
"learning_rate": 4.763499636639389e-05,
|
| 121 |
+
"loss": 4.6213,
|
| 122 |
+
"step": 1500
|
| 123 |
+
},
|
| 124 |
+
{
|
| 125 |
+
"epoch": 0.05055356149840756,
|
| 126 |
+
"grad_norm": 5.697903633117676,
|
| 127 |
+
"learning_rate": 4.74770134917375e-05,
|
| 128 |
+
"loss": 4.6791,
|
| 129 |
+
"step": 1600
|
| 130 |
+
},
|
| 131 |
+
{
|
| 132 |
+
"epoch": 0.05371315909205804,
|
| 133 |
+
"grad_norm": 4.182404041290283,
|
| 134 |
+
"learning_rate": 4.7319030617081114e-05,
|
| 135 |
+
"loss": 4.6493,
|
| 136 |
+
"step": 1700
|
| 137 |
+
},
|
| 138 |
+
{
|
| 139 |
+
"epoch": 0.05687275668570851,
|
| 140 |
+
"grad_norm": 3.8856332302093506,
|
| 141 |
+
"learning_rate": 4.7161047742424724e-05,
|
| 142 |
+
"loss": 4.6081,
|
| 143 |
+
"step": 1800
|
| 144 |
+
},
|
| 145 |
+
{
|
| 146 |
+
"epoch": 0.06003235427935898,
|
| 147 |
+
"grad_norm": 3.778503656387329,
|
| 148 |
+
"learning_rate": 4.700306486776834e-05,
|
| 149 |
+
"loss": 4.6437,
|
| 150 |
+
"step": 1900
|
| 151 |
+
},
|
| 152 |
+
{
|
| 153 |
+
"epoch": 0.06319195187300945,
|
| 154 |
+
"grad_norm": 5.072572231292725,
|
| 155 |
+
"learning_rate": 4.684508199311195e-05,
|
| 156 |
+
"loss": 4.5587,
|
| 157 |
+
"step": 2000
|
| 158 |
+
},
|
| 159 |
+
{
|
| 160 |
+
"epoch": 0.06319195187300945,
|
| 161 |
+
"eval_loss": 4.763667106628418,
|
| 162 |
+
"eval_runtime": 665.5888,
|
| 163 |
+
"eval_samples_per_second": 132.088,
|
| 164 |
+
"eval_steps_per_second": 16.512,
|
| 165 |
+
"step": 2000
|
| 166 |
+
},
|
| 167 |
+
{
|
| 168 |
+
"epoch": 0.06635154946665993,
|
| 169 |
+
"grad_norm": 5.395784854888916,
|
| 170 |
+
"learning_rate": 4.6687099118455565e-05,
|
| 171 |
+
"loss": 4.5621,
|
| 172 |
+
"step": 2100
|
| 173 |
+
},
|
| 174 |
+
{
|
| 175 |
+
"epoch": 0.0695111470603104,
|
| 176 |
+
"grad_norm": 4.856666088104248,
|
| 177 |
+
"learning_rate": 4.6529116243799174e-05,
|
| 178 |
+
"loss": 4.5721,
|
| 179 |
+
"step": 2200
|
| 180 |
+
},
|
| 181 |
+
{
|
| 182 |
+
"epoch": 0.07267074465396087,
|
| 183 |
+
"grad_norm": 4.935929298400879,
|
| 184 |
+
"learning_rate": 4.637113336914279e-05,
|
| 185 |
+
"loss": 4.6004,
|
| 186 |
+
"step": 2300
|
| 187 |
+
},
|
| 188 |
+
{
|
| 189 |
+
"epoch": 0.07583034224761134,
|
| 190 |
+
"grad_norm": 5.257358551025391,
|
| 191 |
+
"learning_rate": 4.62131504944864e-05,
|
| 192 |
+
"loss": 4.6032,
|
| 193 |
+
"step": 2400
|
| 194 |
+
},
|
| 195 |
+
{
|
| 196 |
+
"epoch": 0.07898993984126182,
|
| 197 |
+
"grad_norm": 4.67532205581665,
|
| 198 |
+
"learning_rate": 4.6055167619830016e-05,
|
| 199 |
+
"loss": 4.5734,
|
| 200 |
+
"step": 2500
|
| 201 |
+
},
|
| 202 |
+
{
|
| 203 |
+
"epoch": 0.0821495374349123,
|
| 204 |
+
"grad_norm": 4.023921966552734,
|
| 205 |
+
"learning_rate": 4.5897184745173625e-05,
|
| 206 |
+
"loss": 4.5888,
|
| 207 |
+
"step": 2600
|
| 208 |
+
},
|
| 209 |
+
{
|
| 210 |
+
"epoch": 0.08530913502856276,
|
| 211 |
+
"grad_norm": 3.9619414806365967,
|
| 212 |
+
"learning_rate": 4.573920187051724e-05,
|
| 213 |
+
"loss": 4.5703,
|
| 214 |
+
"step": 2700
|
| 215 |
+
},
|
| 216 |
+
{
|
| 217 |
+
"epoch": 0.08846873262221323,
|
| 218 |
+
"grad_norm": 3.7340047359466553,
|
| 219 |
+
"learning_rate": 4.558121899586085e-05,
|
| 220 |
+
"loss": 4.5437,
|
| 221 |
+
"step": 2800
|
| 222 |
+
},
|
| 223 |
+
{
|
| 224 |
+
"epoch": 0.09162833021586371,
|
| 225 |
+
"grad_norm": 3.5358519554138184,
|
| 226 |
+
"learning_rate": 4.5423236121204466e-05,
|
| 227 |
+
"loss": 4.5433,
|
| 228 |
+
"step": 2900
|
| 229 |
+
},
|
| 230 |
+
{
|
| 231 |
+
"epoch": 0.09478792780951419,
|
| 232 |
+
"grad_norm": 5.275386810302734,
|
| 233 |
+
"learning_rate": 4.5265253246548075e-05,
|
| 234 |
+
"loss": 4.5315,
|
| 235 |
+
"step": 3000
|
| 236 |
+
},
|
| 237 |
+
{
|
| 238 |
+
"epoch": 0.09478792780951419,
|
| 239 |
+
"eval_loss": 4.6958231925964355,
|
| 240 |
+
"eval_runtime": 660.1374,
|
| 241 |
+
"eval_samples_per_second": 133.178,
|
| 242 |
+
"eval_steps_per_second": 16.648,
|
| 243 |
+
"step": 3000
|
| 244 |
+
},
|
| 245 |
+
{
|
| 246 |
+
"epoch": 0.09794752540316465,
|
| 247 |
+
"grad_norm": 4.195945739746094,
|
| 248 |
+
"learning_rate": 4.510727037189169e-05,
|
| 249 |
+
"loss": 4.5292,
|
| 250 |
+
"step": 3100
|
| 251 |
+
},
|
| 252 |
+
{
|
| 253 |
+
"epoch": 0.10110712299681512,
|
| 254 |
+
"grad_norm": 4.976147651672363,
|
| 255 |
+
"learning_rate": 4.49492874972353e-05,
|
| 256 |
+
"loss": 4.5718,
|
| 257 |
+
"step": 3200
|
| 258 |
+
},
|
| 259 |
+
{
|
| 260 |
+
"epoch": 0.1042667205904656,
|
| 261 |
+
"grad_norm": 3.999481678009033,
|
| 262 |
+
"learning_rate": 4.479130462257892e-05,
|
| 263 |
+
"loss": 4.5177,
|
| 264 |
+
"step": 3300
|
| 265 |
+
},
|
| 266 |
+
{
|
| 267 |
+
"epoch": 0.10742631818411608,
|
| 268 |
+
"grad_norm": 5.00895357131958,
|
| 269 |
+
"learning_rate": 4.463332174792253e-05,
|
| 270 |
+
"loss": 4.5418,
|
| 271 |
+
"step": 3400
|
| 272 |
+
},
|
| 273 |
+
{
|
| 274 |
+
"epoch": 0.11058591577776654,
|
| 275 |
+
"grad_norm": 4.499265193939209,
|
| 276 |
+
"learning_rate": 4.447533887326614e-05,
|
| 277 |
+
"loss": 4.5262,
|
| 278 |
+
"step": 3500
|
| 279 |
+
},
|
| 280 |
+
{
|
| 281 |
+
"epoch": 0.11374551337141701,
|
| 282 |
+
"grad_norm": 4.686629772186279,
|
| 283 |
+
"learning_rate": 4.431735599860976e-05,
|
| 284 |
+
"loss": 4.5276,
|
| 285 |
+
"step": 3600
|
| 286 |
+
},
|
| 287 |
+
{
|
| 288 |
+
"epoch": 0.11690511096506749,
|
| 289 |
+
"grad_norm": 5.327017784118652,
|
| 290 |
+
"learning_rate": 4.415937312395337e-05,
|
| 291 |
+
"loss": 4.4939,
|
| 292 |
+
"step": 3700
|
| 293 |
+
},
|
| 294 |
+
{
|
| 295 |
+
"epoch": 0.12006470855871797,
|
| 296 |
+
"grad_norm": 3.8358302116394043,
|
| 297 |
+
"learning_rate": 4.400139024929698e-05,
|
| 298 |
+
"loss": 4.5028,
|
| 299 |
+
"step": 3800
|
| 300 |
+
},
|
| 301 |
+
{
|
| 302 |
+
"epoch": 0.12322430615236843,
|
| 303 |
+
"grad_norm": 4.16530704498291,
|
| 304 |
+
"learning_rate": 4.384340737464059e-05,
|
| 305 |
+
"loss": 4.4964,
|
| 306 |
+
"step": 3900
|
| 307 |
+
},
|
| 308 |
+
{
|
| 309 |
+
"epoch": 0.1263839037460189,
|
| 310 |
+
"grad_norm": 4.142209053039551,
|
| 311 |
+
"learning_rate": 4.36854244999842e-05,
|
| 312 |
+
"loss": 4.4899,
|
| 313 |
+
"step": 4000
|
| 314 |
+
},
|
| 315 |
+
{
|
| 316 |
+
"epoch": 0.1263839037460189,
|
| 317 |
+
"eval_loss": 4.649657249450684,
|
| 318 |
+
"eval_runtime": 689.9636,
|
| 319 |
+
"eval_samples_per_second": 127.421,
|
| 320 |
+
"eval_steps_per_second": 15.928,
|
| 321 |
+
"step": 4000
|
| 322 |
+
},
|
| 323 |
+
{
|
| 324 |
+
"epoch": 0.12954350133966938,
|
| 325 |
+
"grad_norm": 3.615328311920166,
|
| 326 |
+
"learning_rate": 4.352744162532781e-05,
|
| 327 |
+
"loss": 4.4671,
|
| 328 |
+
"step": 4100
|
| 329 |
+
},
|
| 330 |
+
{
|
| 331 |
+
"epoch": 0.13270309893331986,
|
| 332 |
+
"grad_norm": 5.139493465423584,
|
| 333 |
+
"learning_rate": 4.336945875067143e-05,
|
| 334 |
+
"loss": 4.4954,
|
| 335 |
+
"step": 4200
|
| 336 |
+
},
|
| 337 |
+
{
|
| 338 |
+
"epoch": 0.13586269652697033,
|
| 339 |
+
"grad_norm": 4.381932258605957,
|
| 340 |
+
"learning_rate": 4.3211475876015036e-05,
|
| 341 |
+
"loss": 4.5103,
|
| 342 |
+
"step": 4300
|
| 343 |
+
},
|
| 344 |
+
{
|
| 345 |
+
"epoch": 0.1390222941206208,
|
| 346 |
+
"grad_norm": 4.257575988769531,
|
| 347 |
+
"learning_rate": 4.305349300135865e-05,
|
| 348 |
+
"loss": 4.4677,
|
| 349 |
+
"step": 4400
|
| 350 |
+
},
|
| 351 |
+
{
|
| 352 |
+
"epoch": 0.14218189171427126,
|
| 353 |
+
"grad_norm": 3.5805318355560303,
|
| 354 |
+
"learning_rate": 4.289551012670227e-05,
|
| 355 |
+
"loss": 4.4547,
|
| 356 |
+
"step": 4500
|
| 357 |
+
},
|
| 358 |
+
{
|
| 359 |
+
"epoch": 0.14534148930792173,
|
| 360 |
+
"grad_norm": 4.602927207946777,
|
| 361 |
+
"learning_rate": 4.273752725204588e-05,
|
| 362 |
+
"loss": 4.4885,
|
| 363 |
+
"step": 4600
|
| 364 |
+
},
|
| 365 |
+
{
|
| 366 |
+
"epoch": 0.1485010869015722,
|
| 367 |
+
"grad_norm": 4.054294586181641,
|
| 368 |
+
"learning_rate": 4.2579544377389494e-05,
|
| 369 |
+
"loss": 4.4791,
|
| 370 |
+
"step": 4700
|
| 371 |
+
},
|
| 372 |
+
{
|
| 373 |
+
"epoch": 0.1516606844952227,
|
| 374 |
+
"grad_norm": 3.857389450073242,
|
| 375 |
+
"learning_rate": 4.24215615027331e-05,
|
| 376 |
+
"loss": 4.4316,
|
| 377 |
+
"step": 4800
|
| 378 |
+
},
|
| 379 |
+
{
|
| 380 |
+
"epoch": 0.15482028208887316,
|
| 381 |
+
"grad_norm": 3.913726806640625,
|
| 382 |
+
"learning_rate": 4.226357862807672e-05,
|
| 383 |
+
"loss": 4.4786,
|
| 384 |
+
"step": 4900
|
| 385 |
+
},
|
| 386 |
+
{
|
| 387 |
+
"epoch": 0.15797987968252364,
|
| 388 |
+
"grad_norm": 4.159733295440674,
|
| 389 |
+
"learning_rate": 4.210559575342033e-05,
|
| 390 |
+
"loss": 4.422,
|
| 391 |
+
"step": 5000
|
| 392 |
+
},
|
| 393 |
+
{
|
| 394 |
+
"epoch": 0.15797987968252364,
|
| 395 |
+
"eval_loss": 4.6132493019104,
|
| 396 |
+
"eval_runtime": 721.543,
|
| 397 |
+
"eval_samples_per_second": 121.844,
|
| 398 |
+
"eval_steps_per_second": 15.231,
|
| 399 |
+
"step": 5000
|
| 400 |
+
},
|
| 401 |
+
{
|
| 402 |
+
"epoch": 0.16113947727617411,
|
| 403 |
+
"grad_norm": 5.403946876525879,
|
| 404 |
+
"learning_rate": 4.1947612878763944e-05,
|
| 405 |
+
"loss": 4.4857,
|
| 406 |
+
"step": 5100
|
| 407 |
+
},
|
| 408 |
+
{
|
| 409 |
+
"epoch": 0.1642990748698246,
|
| 410 |
+
"grad_norm": 11.68526554107666,
|
| 411 |
+
"learning_rate": 4.178963000410755e-05,
|
| 412 |
+
"loss": 4.4777,
|
| 413 |
+
"step": 5200
|
| 414 |
+
},
|
| 415 |
+
{
|
| 416 |
+
"epoch": 0.16745867246347504,
|
| 417 |
+
"grad_norm": 4.937869548797607,
|
| 418 |
+
"learning_rate": 4.163164712945117e-05,
|
| 419 |
+
"loss": 4.4452,
|
| 420 |
+
"step": 5300
|
| 421 |
+
},
|
| 422 |
+
{
|
| 423 |
+
"epoch": 0.17061827005712552,
|
| 424 |
+
"grad_norm": 3.7432169914245605,
|
| 425 |
+
"learning_rate": 4.147366425479478e-05,
|
| 426 |
+
"loss": 4.4717,
|
| 427 |
+
"step": 5400
|
| 428 |
+
},
|
| 429 |
+
{
|
| 430 |
+
"epoch": 0.173777867650776,
|
| 431 |
+
"grad_norm": 3.801788806915283,
|
| 432 |
+
"learning_rate": 4.1317261208884964e-05,
|
| 433 |
+
"loss": 4.4424,
|
| 434 |
+
"step": 5500
|
| 435 |
+
},
|
| 436 |
+
{
|
| 437 |
+
"epoch": 0.17693746524442647,
|
| 438 |
+
"grad_norm": 4.984846115112305,
|
| 439 |
+
"learning_rate": 4.115927833422857e-05,
|
| 440 |
+
"loss": 4.4671,
|
| 441 |
+
"step": 5600
|
| 442 |
+
},
|
| 443 |
+
{
|
| 444 |
+
"epoch": 0.18009706283807694,
|
| 445 |
+
"grad_norm": 3.8125479221343994,
|
| 446 |
+
"learning_rate": 4.100287528831875e-05,
|
| 447 |
+
"loss": 4.4079,
|
| 448 |
+
"step": 5700
|
| 449 |
+
},
|
| 450 |
+
{
|
| 451 |
+
"epoch": 0.18325666043172742,
|
| 452 |
+
"grad_norm": 5.528467655181885,
|
| 453 |
+
"learning_rate": 4.084489241366236e-05,
|
| 454 |
+
"loss": 4.4194,
|
| 455 |
+
"step": 5800
|
| 456 |
+
},
|
| 457 |
+
{
|
| 458 |
+
"epoch": 0.1864162580253779,
|
| 459 |
+
"grad_norm": 4.0301899909973145,
|
| 460 |
+
"learning_rate": 4.068690953900598e-05,
|
| 461 |
+
"loss": 4.4299,
|
| 462 |
+
"step": 5900
|
| 463 |
+
},
|
| 464 |
+
{
|
| 465 |
+
"epoch": 0.18957585561902837,
|
| 466 |
+
"grad_norm": 3.4992246627807617,
|
| 467 |
+
"learning_rate": 4.0528926664349586e-05,
|
| 468 |
+
"loss": 4.448,
|
| 469 |
+
"step": 6000
|
| 470 |
+
},
|
| 471 |
+
{
|
| 472 |
+
"epoch": 0.18957585561902837,
|
| 473 |
+
"eval_loss": 4.5857086181640625,
|
| 474 |
+
"eval_runtime": 702.1155,
|
| 475 |
+
"eval_samples_per_second": 125.216,
|
| 476 |
+
"eval_steps_per_second": 15.653,
|
| 477 |
+
"step": 6000
|
| 478 |
+
},
|
| 479 |
+
{
|
| 480 |
+
"epoch": 0.19273545321267882,
|
| 481 |
+
"grad_norm": 4.121975421905518,
|
| 482 |
+
"learning_rate": 4.03709437896932e-05,
|
| 483 |
+
"loss": 4.4076,
|
| 484 |
+
"step": 6100
|
| 485 |
+
},
|
| 486 |
+
{
|
| 487 |
+
"epoch": 0.1958950508063293,
|
| 488 |
+
"grad_norm": 3.662929058074951,
|
| 489 |
+
"learning_rate": 4.021296091503681e-05,
|
| 490 |
+
"loss": 4.3953,
|
| 491 |
+
"step": 6200
|
| 492 |
+
},
|
| 493 |
+
{
|
| 494 |
+
"epoch": 0.19905464839997977,
|
| 495 |
+
"grad_norm": 4.008944988250732,
|
| 496 |
+
"learning_rate": 4.005497804038043e-05,
|
| 497 |
+
"loss": 4.4581,
|
| 498 |
+
"step": 6300
|
| 499 |
+
},
|
| 500 |
+
{
|
| 501 |
+
"epoch": 0.20221424599363025,
|
| 502 |
+
"grad_norm": 3.7961764335632324,
|
| 503 |
+
"learning_rate": 3.9896995165724036e-05,
|
| 504 |
+
"loss": 4.397,
|
| 505 |
+
"step": 6400
|
| 506 |
+
},
|
| 507 |
+
{
|
| 508 |
+
"epoch": 0.20537384358728072,
|
| 509 |
+
"grad_norm": 4.710026741027832,
|
| 510 |
+
"learning_rate": 3.973901229106765e-05,
|
| 511 |
+
"loss": 4.4515,
|
| 512 |
+
"step": 6500
|
| 513 |
+
},
|
| 514 |
+
{
|
| 515 |
+
"epoch": 0.2085334411809312,
|
| 516 |
+
"grad_norm": 4.220251083374023,
|
| 517 |
+
"learning_rate": 3.958102941641126e-05,
|
| 518 |
+
"loss": 4.4199,
|
| 519 |
+
"step": 6600
|
| 520 |
+
},
|
| 521 |
+
{
|
| 522 |
+
"epoch": 0.21169303877458168,
|
| 523 |
+
"grad_norm": 3.7243406772613525,
|
| 524 |
+
"learning_rate": 3.942304654175488e-05,
|
| 525 |
+
"loss": 4.4097,
|
| 526 |
+
"step": 6700
|
| 527 |
+
},
|
| 528 |
+
{
|
| 529 |
+
"epoch": 0.21485263636823215,
|
| 530 |
+
"grad_norm": 4.340923309326172,
|
| 531 |
+
"learning_rate": 3.926664349584505e-05,
|
| 532 |
+
"loss": 4.4105,
|
| 533 |
+
"step": 6800
|
| 534 |
+
},
|
| 535 |
+
{
|
| 536 |
+
"epoch": 0.2180122339618826,
|
| 537 |
+
"grad_norm": 4.304684638977051,
|
| 538 |
+
"learning_rate": 3.9108660621188665e-05,
|
| 539 |
+
"loss": 4.4033,
|
| 540 |
+
"step": 6900
|
| 541 |
+
},
|
| 542 |
+
{
|
| 543 |
+
"epoch": 0.22117183155553308,
|
| 544 |
+
"grad_norm": 4.587896347045898,
|
| 545 |
+
"learning_rate": 3.8950677746532275e-05,
|
| 546 |
+
"loss": 4.4451,
|
| 547 |
+
"step": 7000
|
| 548 |
+
},
|
| 549 |
+
{
|
| 550 |
+
"epoch": 0.22117183155553308,
|
| 551 |
+
"eval_loss": 4.5615153312683105,
|
| 552 |
+
"eval_runtime": 695.974,
|
| 553 |
+
"eval_samples_per_second": 126.321,
|
| 554 |
+
"eval_steps_per_second": 15.791,
|
| 555 |
+
"step": 7000
|
| 556 |
+
},
|
| 557 |
+
{
|
| 558 |
+
"epoch": 0.22433142914918355,
|
| 559 |
+
"grad_norm": 4.658904075622559,
|
| 560 |
+
"learning_rate": 3.879269487187589e-05,
|
| 561 |
+
"loss": 4.407,
|
| 562 |
+
"step": 7100
|
| 563 |
+
},
|
| 564 |
+
{
|
| 565 |
+
"epoch": 0.22749102674283403,
|
| 566 |
+
"grad_norm": 4.226349830627441,
|
| 567 |
+
"learning_rate": 3.86347119972195e-05,
|
| 568 |
+
"loss": 4.4055,
|
| 569 |
+
"step": 7200
|
| 570 |
+
},
|
| 571 |
+
{
|
| 572 |
+
"epoch": 0.2306506243364845,
|
| 573 |
+
"grad_norm": 5.705229759216309,
|
| 574 |
+
"learning_rate": 3.8476729122563116e-05,
|
| 575 |
+
"loss": 4.3862,
|
| 576 |
+
"step": 7300
|
| 577 |
+
},
|
| 578 |
+
{
|
| 579 |
+
"epoch": 0.23381022193013498,
|
| 580 |
+
"grad_norm": 3.9644031524658203,
|
| 581 |
+
"learning_rate": 3.8318746247906725e-05,
|
| 582 |
+
"loss": 4.3647,
|
| 583 |
+
"step": 7400
|
| 584 |
+
},
|
| 585 |
+
{
|
| 586 |
+
"epoch": 0.23696981952378546,
|
| 587 |
+
"grad_norm": 4.256899356842041,
|
| 588 |
+
"learning_rate": 3.816076337325034e-05,
|
| 589 |
+
"loss": 4.4091,
|
| 590 |
+
"step": 7500
|
| 591 |
+
},
|
| 592 |
+
{
|
| 593 |
+
"epoch": 0.24012941711743593,
|
| 594 |
+
"grad_norm": 3.951305627822876,
|
| 595 |
+
"learning_rate": 3.800278049859396e-05,
|
| 596 |
+
"loss": 4.3994,
|
| 597 |
+
"step": 7600
|
| 598 |
+
},
|
| 599 |
+
{
|
| 600 |
+
"epoch": 0.2432890147110864,
|
| 601 |
+
"grad_norm": 3.6965837478637695,
|
| 602 |
+
"learning_rate": 3.7844797623937566e-05,
|
| 603 |
+
"loss": 4.4043,
|
| 604 |
+
"step": 7700
|
| 605 |
+
},
|
| 606 |
+
{
|
| 607 |
+
"epoch": 0.24644861230473686,
|
| 608 |
+
"grad_norm": 4.412446022033691,
|
| 609 |
+
"learning_rate": 3.768681474928118e-05,
|
| 610 |
+
"loss": 4.4237,
|
| 611 |
+
"step": 7800
|
| 612 |
+
},
|
| 613 |
+
{
|
| 614 |
+
"epoch": 0.24960820989838733,
|
| 615 |
+
"grad_norm": 4.190601348876953,
|
| 616 |
+
"learning_rate": 3.752883187462479e-05,
|
| 617 |
+
"loss": 4.365,
|
| 618 |
+
"step": 7900
|
| 619 |
+
},
|
| 620 |
+
{
|
| 621 |
+
"epoch": 0.2527678074920378,
|
| 622 |
+
"grad_norm": 3.6600217819213867,
|
| 623 |
+
"learning_rate": 3.737084899996841e-05,
|
| 624 |
+
"loss": 4.393,
|
| 625 |
+
"step": 8000
|
| 626 |
+
},
|
| 627 |
+
{
|
| 628 |
+
"epoch": 0.2527678074920378,
|
| 629 |
+
"eval_loss": 4.542718887329102,
|
| 630 |
+
"eval_runtime": 692.645,
|
| 631 |
+
"eval_samples_per_second": 126.928,
|
| 632 |
+
"eval_steps_per_second": 15.867,
|
| 633 |
+
"step": 8000
|
| 634 |
+
},
|
| 635 |
+
{
|
| 636 |
+
"epoch": 0.2559274050856883,
|
| 637 |
+
"grad_norm": 6.242320537567139,
|
| 638 |
+
"learning_rate": 3.721286612531202e-05,
|
| 639 |
+
"loss": 4.4185,
|
| 640 |
+
"step": 8100
|
| 641 |
+
},
|
| 642 |
+
{
|
| 643 |
+
"epoch": 0.25908700267933876,
|
| 644 |
+
"grad_norm": 4.762045383453369,
|
| 645 |
+
"learning_rate": 3.705488325065563e-05,
|
| 646 |
+
"loss": 4.3656,
|
| 647 |
+
"step": 8200
|
| 648 |
+
},
|
| 649 |
+
{
|
| 650 |
+
"epoch": 0.26224660027298924,
|
| 651 |
+
"grad_norm": 5.066886901855469,
|
| 652 |
+
"learning_rate": 3.689690037599924e-05,
|
| 653 |
+
"loss": 4.394,
|
| 654 |
+
"step": 8300
|
| 655 |
+
},
|
| 656 |
+
{
|
| 657 |
+
"epoch": 0.2654061978666397,
|
| 658 |
+
"grad_norm": 4.975124835968018,
|
| 659 |
+
"learning_rate": 3.673891750134286e-05,
|
| 660 |
+
"loss": 4.3346,
|
| 661 |
+
"step": 8400
|
| 662 |
+
},
|
| 663 |
+
{
|
| 664 |
+
"epoch": 0.2685657954602902,
|
| 665 |
+
"grad_norm": 4.436885356903076,
|
| 666 |
+
"learning_rate": 3.658093462668647e-05,
|
| 667 |
+
"loss": 4.3958,
|
| 668 |
+
"step": 8500
|
| 669 |
+
},
|
| 670 |
+
{
|
| 671 |
+
"epoch": 0.27172539305394067,
|
| 672 |
+
"grad_norm": 4.618080139160156,
|
| 673 |
+
"learning_rate": 3.6422951752030084e-05,
|
| 674 |
+
"loss": 4.3791,
|
| 675 |
+
"step": 8600
|
| 676 |
+
},
|
| 677 |
+
{
|
| 678 |
+
"epoch": 0.27488499064759114,
|
| 679 |
+
"grad_norm": 4.878961086273193,
|
| 680 |
+
"learning_rate": 3.626496887737369e-05,
|
| 681 |
+
"loss": 4.3738,
|
| 682 |
+
"step": 8700
|
| 683 |
+
},
|
| 684 |
+
{
|
| 685 |
+
"epoch": 0.2780445882412416,
|
| 686 |
+
"grad_norm": 3.7712392807006836,
|
| 687 |
+
"learning_rate": 3.610698600271731e-05,
|
| 688 |
+
"loss": 4.362,
|
| 689 |
+
"step": 8800
|
| 690 |
+
},
|
| 691 |
+
{
|
| 692 |
+
"epoch": 0.2812041858348921,
|
| 693 |
+
"grad_norm": 4.318520545959473,
|
| 694 |
+
"learning_rate": 3.594900312806092e-05,
|
| 695 |
+
"loss": 4.367,
|
| 696 |
+
"step": 8900
|
| 697 |
+
},
|
| 698 |
+
{
|
| 699 |
+
"epoch": 0.2843637834285425,
|
| 700 |
+
"grad_norm": 3.3742518424987793,
|
| 701 |
+
"learning_rate": 3.5791020253404534e-05,
|
| 702 |
+
"loss": 4.3763,
|
| 703 |
+
"step": 9000
|
| 704 |
+
},
|
| 705 |
+
{
|
| 706 |
+
"epoch": 0.2843637834285425,
|
| 707 |
+
"eval_loss": 4.525923728942871,
|
| 708 |
+
"eval_runtime": 692.5212,
|
| 709 |
+
"eval_samples_per_second": 126.951,
|
| 710 |
+
"eval_steps_per_second": 15.87,
|
| 711 |
+
"step": 9000
|
| 712 |
+
},
|
| 713 |
+
{
|
| 714 |
+
"epoch": 0.287523381022193,
|
| 715 |
+
"grad_norm": 4.027231693267822,
|
| 716 |
+
"learning_rate": 3.5633037378748143e-05,
|
| 717 |
+
"loss": 4.4086,
|
| 718 |
+
"step": 9100
|
| 719 |
+
},
|
| 720 |
+
{
|
| 721 |
+
"epoch": 0.29068297861584347,
|
| 722 |
+
"grad_norm": 4.556132793426514,
|
| 723 |
+
"learning_rate": 3.547505450409176e-05,
|
| 724 |
+
"loss": 4.3686,
|
| 725 |
+
"step": 9200
|
| 726 |
+
},
|
| 727 |
+
{
|
| 728 |
+
"epoch": 0.29384257620949394,
|
| 729 |
+
"grad_norm": 3.9203240871429443,
|
| 730 |
+
"learning_rate": 3.531707162943537e-05,
|
| 731 |
+
"loss": 4.365,
|
| 732 |
+
"step": 9300
|
| 733 |
+
},
|
| 734 |
+
{
|
| 735 |
+
"epoch": 0.2970021738031444,
|
| 736 |
+
"grad_norm": 4.094367027282715,
|
| 737 |
+
"learning_rate": 3.5159088754778985e-05,
|
| 738 |
+
"loss": 4.3734,
|
| 739 |
+
"step": 9400
|
| 740 |
+
},
|
| 741 |
+
{
|
| 742 |
+
"epoch": 0.3001617713967949,
|
| 743 |
+
"grad_norm": 5.018619060516357,
|
| 744 |
+
"learning_rate": 3.50011058801226e-05,
|
| 745 |
+
"loss": 4.4014,
|
| 746 |
+
"step": 9500
|
| 747 |
+
},
|
| 748 |
+
{
|
| 749 |
+
"epoch": 0.3033213689904454,
|
| 750 |
+
"grad_norm": 3.928473711013794,
|
| 751 |
+
"learning_rate": 3.484312300546621e-05,
|
| 752 |
+
"loss": 4.3508,
|
| 753 |
+
"step": 9600
|
| 754 |
+
},
|
| 755 |
+
{
|
| 756 |
+
"epoch": 0.30648096658409585,
|
| 757 |
+
"grad_norm": 3.889299154281616,
|
| 758 |
+
"learning_rate": 3.4685140130809826e-05,
|
| 759 |
+
"loss": 4.365,
|
| 760 |
+
"step": 9700
|
| 761 |
+
},
|
| 762 |
+
{
|
| 763 |
+
"epoch": 0.3096405641777463,
|
| 764 |
+
"grad_norm": 4.932437896728516,
|
| 765 |
+
"learning_rate": 3.4527157256153435e-05,
|
| 766 |
+
"loss": 4.3166,
|
| 767 |
+
"step": 9800
|
| 768 |
+
},
|
| 769 |
+
{
|
| 770 |
+
"epoch": 0.3128001617713968,
|
| 771 |
+
"grad_norm": 4.175967216491699,
|
| 772 |
+
"learning_rate": 3.436917438149705e-05,
|
| 773 |
+
"loss": 4.3463,
|
| 774 |
+
"step": 9900
|
| 775 |
+
},
|
| 776 |
+
{
|
| 777 |
+
"epoch": 0.3159597593650473,
|
| 778 |
+
"grad_norm": 5.54374361038208,
|
| 779 |
+
"learning_rate": 3.421119150684066e-05,
|
| 780 |
+
"loss": 4.3506,
|
| 781 |
+
"step": 10000
|
| 782 |
+
},
|
| 783 |
+
{
|
| 784 |
+
"epoch": 0.3159597593650473,
|
| 785 |
+
"eval_loss": 4.511664390563965,
|
| 786 |
+
"eval_runtime": 691.9773,
|
| 787 |
+
"eval_samples_per_second": 127.05,
|
| 788 |
+
"eval_steps_per_second": 15.882,
|
| 789 |
+
"step": 10000
|
| 790 |
+
},
|
| 791 |
+
{
|
| 792 |
+
"epoch": 0.31911935695869775,
|
| 793 |
+
"grad_norm": 4.466663360595703,
|
| 794 |
+
"learning_rate": 3.4053208632184277e-05,
|
| 795 |
+
"loss": 4.3351,
|
| 796 |
+
"step": 10100
|
| 797 |
+
},
|
| 798 |
+
{
|
| 799 |
+
"epoch": 0.32227895455234823,
|
| 800 |
+
"grad_norm": 4.300159454345703,
|
| 801 |
+
"learning_rate": 3.3895225757527886e-05,
|
| 802 |
+
"loss": 4.3605,
|
| 803 |
+
"step": 10200
|
| 804 |
+
},
|
| 805 |
+
{
|
| 806 |
+
"epoch": 0.3254385521459987,
|
| 807 |
+
"grad_norm": 4.595406532287598,
|
| 808 |
+
"learning_rate": 3.37372428828715e-05,
|
| 809 |
+
"loss": 4.3432,
|
| 810 |
+
"step": 10300
|
| 811 |
+
},
|
| 812 |
+
{
|
| 813 |
+
"epoch": 0.3285981497396492,
|
| 814 |
+
"grad_norm": 4.077473163604736,
|
| 815 |
+
"learning_rate": 3.357926000821511e-05,
|
| 816 |
+
"loss": 4.3421,
|
| 817 |
+
"step": 10400
|
| 818 |
+
},
|
| 819 |
+
{
|
| 820 |
+
"epoch": 0.33175774733329966,
|
| 821 |
+
"grad_norm": 4.250525951385498,
|
| 822 |
+
"learning_rate": 3.342127713355873e-05,
|
| 823 |
+
"loss": 4.3795,
|
| 824 |
+
"step": 10500
|
| 825 |
+
},
|
| 826 |
+
{
|
| 827 |
+
"epoch": 0.3349173449269501,
|
| 828 |
+
"grad_norm": 3.9417407512664795,
|
| 829 |
+
"learning_rate": 3.3263294258902336e-05,
|
| 830 |
+
"loss": 4.3394,
|
| 831 |
+
"step": 10600
|
| 832 |
+
},
|
| 833 |
+
{
|
| 834 |
+
"epoch": 0.33807694252060055,
|
| 835 |
+
"grad_norm": 3.782409906387329,
|
| 836 |
+
"learning_rate": 3.310531138424595e-05,
|
| 837 |
+
"loss": 4.3061,
|
| 838 |
+
"step": 10700
|
| 839 |
+
},
|
| 840 |
+
{
|
| 841 |
+
"epoch": 0.34123654011425103,
|
| 842 |
+
"grad_norm": 4.497501850128174,
|
| 843 |
+
"learning_rate": 3.294732850958956e-05,
|
| 844 |
+
"loss": 4.3558,
|
| 845 |
+
"step": 10800
|
| 846 |
+
},
|
| 847 |
+
{
|
| 848 |
+
"epoch": 0.3443961377079015,
|
| 849 |
+
"grad_norm": 4.938557147979736,
|
| 850 |
+
"learning_rate": 3.278934563493318e-05,
|
| 851 |
+
"loss": 4.33,
|
| 852 |
+
"step": 10900
|
| 853 |
+
},
|
| 854 |
+
{
|
| 855 |
+
"epoch": 0.347555735301552,
|
| 856 |
+
"grad_norm": 4.101445198059082,
|
| 857 |
+
"learning_rate": 3.263136276027679e-05,
|
| 858 |
+
"loss": 4.3658,
|
| 859 |
+
"step": 11000
|
| 860 |
+
},
|
| 861 |
+
{
|
| 862 |
+
"epoch": 0.347555735301552,
|
| 863 |
+
"eval_loss": 4.499057769775391,
|
| 864 |
+
"eval_runtime": 696.9678,
|
| 865 |
+
"eval_samples_per_second": 126.141,
|
| 866 |
+
"eval_steps_per_second": 15.768,
|
| 867 |
+
"step": 11000
|
| 868 |
+
},
|
| 869 |
+
{
|
| 870 |
+
"epoch": 0.35071533289520246,
|
| 871 |
+
"grad_norm": 4.350194454193115,
|
| 872 |
+
"learning_rate": 3.24733798856204e-05,
|
| 873 |
+
"loss": 4.3129,
|
| 874 |
+
"step": 11100
|
| 875 |
+
},
|
| 876 |
+
{
|
| 877 |
+
"epoch": 0.35387493048885293,
|
| 878 |
+
"grad_norm": 4.244603157043457,
|
| 879 |
+
"learning_rate": 3.231539701096401e-05,
|
| 880 |
+
"loss": 4.3566,
|
| 881 |
+
"step": 11200
|
| 882 |
+
},
|
| 883 |
+
{
|
| 884 |
+
"epoch": 0.3570345280825034,
|
| 885 |
+
"grad_norm": 3.947141647338867,
|
| 886 |
+
"learning_rate": 3.215741413630762e-05,
|
| 887 |
+
"loss": 4.3572,
|
| 888 |
+
"step": 11300
|
| 889 |
+
},
|
| 890 |
+
{
|
| 891 |
+
"epoch": 0.3601941256761539,
|
| 892 |
+
"grad_norm": 4.241100311279297,
|
| 893 |
+
"learning_rate": 3.199943126165124e-05,
|
| 894 |
+
"loss": 4.3341,
|
| 895 |
+
"step": 11400
|
| 896 |
+
},
|
| 897 |
+
{
|
| 898 |
+
"epoch": 0.36335372326980436,
|
| 899 |
+
"grad_norm": 4.596353530883789,
|
| 900 |
+
"learning_rate": 3.184144838699485e-05,
|
| 901 |
+
"loss": 4.3702,
|
| 902 |
+
"step": 11500
|
| 903 |
+
},
|
| 904 |
+
{
|
| 905 |
+
"epoch": 0.36651332086345484,
|
| 906 |
+
"grad_norm": 4.667087554931641,
|
| 907 |
+
"learning_rate": 3.168346551233846e-05,
|
| 908 |
+
"loss": 4.3919,
|
| 909 |
+
"step": 11600
|
| 910 |
+
},
|
| 911 |
+
{
|
| 912 |
+
"epoch": 0.3696729184571053,
|
| 913 |
+
"grad_norm": 4.700224876403809,
|
| 914 |
+
"learning_rate": 3.152548263768207e-05,
|
| 915 |
+
"loss": 4.3622,
|
| 916 |
+
"step": 11700
|
| 917 |
+
},
|
| 918 |
+
{
|
| 919 |
+
"epoch": 0.3728325160507558,
|
| 920 |
+
"grad_norm": 4.907840728759766,
|
| 921 |
+
"learning_rate": 3.136749976302569e-05,
|
| 922 |
+
"loss": 4.3517,
|
| 923 |
+
"step": 11800
|
| 924 |
+
},
|
| 925 |
+
{
|
| 926 |
+
"epoch": 0.37599211364440627,
|
| 927 |
+
"grad_norm": 4.013961315155029,
|
| 928 |
+
"learning_rate": 3.12095168883693e-05,
|
| 929 |
+
"loss": 4.3636,
|
| 930 |
+
"step": 11900
|
| 931 |
+
},
|
| 932 |
+
{
|
| 933 |
+
"epoch": 0.37915171123805674,
|
| 934 |
+
"grad_norm": 3.6914377212524414,
|
| 935 |
+
"learning_rate": 3.105153401371291e-05,
|
| 936 |
+
"loss": 4.3087,
|
| 937 |
+
"step": 12000
|
| 938 |
+
},
|
| 939 |
+
{
|
| 940 |
+
"epoch": 0.37915171123805674,
|
| 941 |
+
"eval_loss": 4.48746395111084,
|
| 942 |
+
"eval_runtime": 715.6219,
|
| 943 |
+
"eval_samples_per_second": 122.853,
|
| 944 |
+
"eval_steps_per_second": 15.357,
|
| 945 |
+
"step": 12000
|
| 946 |
+
},
|
| 947 |
+
{
|
| 948 |
+
"epoch": 0.3823113088317072,
|
| 949 |
+
"grad_norm": 3.8867859840393066,
|
| 950 |
+
"learning_rate": 3.089355113905652e-05,
|
| 951 |
+
"loss": 4.337,
|
| 952 |
+
"step": 12100
|
| 953 |
+
},
|
| 954 |
+
{
|
| 955 |
+
"epoch": 0.38547090642535764,
|
| 956 |
+
"grad_norm": 4.648858070373535,
|
| 957 |
+
"learning_rate": 3.073556826440014e-05,
|
| 958 |
+
"loss": 4.3516,
|
| 959 |
+
"step": 12200
|
| 960 |
+
},
|
| 961 |
+
{
|
| 962 |
+
"epoch": 0.3886305040190081,
|
| 963 |
+
"grad_norm": 4.28641939163208,
|
| 964 |
+
"learning_rate": 3.0577585389743755e-05,
|
| 965 |
+
"loss": 4.3053,
|
| 966 |
+
"step": 12300
|
| 967 |
+
},
|
| 968 |
+
{
|
| 969 |
+
"epoch": 0.3917901016126586,
|
| 970 |
+
"grad_norm": 3.784064292907715,
|
| 971 |
+
"learning_rate": 3.0419602515087364e-05,
|
| 972 |
+
"loss": 4.3373,
|
| 973 |
+
"step": 12400
|
| 974 |
+
},
|
| 975 |
+
{
|
| 976 |
+
"epoch": 0.39494969920630907,
|
| 977 |
+
"grad_norm": 3.9344184398651123,
|
| 978 |
+
"learning_rate": 3.0261619640430976e-05,
|
| 979 |
+
"loss": 4.2861,
|
| 980 |
+
"step": 12500
|
| 981 |
+
},
|
| 982 |
+
{
|
| 983 |
+
"epoch": 0.39810929679995954,
|
| 984 |
+
"grad_norm": 4.234601974487305,
|
| 985 |
+
"learning_rate": 3.010363676577459e-05,
|
| 986 |
+
"loss": 4.3603,
|
| 987 |
+
"step": 12600
|
| 988 |
+
},
|
| 989 |
+
{
|
| 990 |
+
"epoch": 0.40126889439361,
|
| 991 |
+
"grad_norm": 4.8835835456848145,
|
| 992 |
+
"learning_rate": 2.9945653891118202e-05,
|
| 993 |
+
"loss": 4.3435,
|
| 994 |
+
"step": 12700
|
| 995 |
+
},
|
| 996 |
+
{
|
| 997 |
+
"epoch": 0.4044284919872605,
|
| 998 |
+
"grad_norm": 4.3561882972717285,
|
| 999 |
+
"learning_rate": 2.9787671016461814e-05,
|
| 1000 |
+
"loss": 4.3272,
|
| 1001 |
+
"step": 12800
|
| 1002 |
+
},
|
| 1003 |
+
{
|
| 1004 |
+
"epoch": 0.407588089580911,
|
| 1005 |
+
"grad_norm": 4.034718036651611,
|
| 1006 |
+
"learning_rate": 2.9629688141805427e-05,
|
| 1007 |
+
"loss": 4.2928,
|
| 1008 |
+
"step": 12900
|
| 1009 |
+
},
|
| 1010 |
+
{
|
| 1011 |
+
"epoch": 0.41074768717456145,
|
| 1012 |
+
"grad_norm": 4.587557315826416,
|
| 1013 |
+
"learning_rate": 2.947170526714904e-05,
|
| 1014 |
+
"loss": 4.3284,
|
| 1015 |
+
"step": 13000
|
| 1016 |
+
},
|
| 1017 |
+
{
|
| 1018 |
+
"epoch": 0.41074768717456145,
|
| 1019 |
+
"eval_loss": 4.477302074432373,
|
| 1020 |
+
"eval_runtime": 710.4818,
|
| 1021 |
+
"eval_samples_per_second": 123.741,
|
| 1022 |
+
"eval_steps_per_second": 15.468,
|
| 1023 |
+
"step": 13000
|
| 1024 |
+
},
|
| 1025 |
+
{
|
| 1026 |
+
"epoch": 0.4139072847682119,
|
| 1027 |
+
"grad_norm": 3.6614646911621094,
|
| 1028 |
+
"learning_rate": 2.9313722392492656e-05,
|
| 1029 |
+
"loss": 4.3301,
|
| 1030 |
+
"step": 13100
|
| 1031 |
+
},
|
| 1032 |
+
{
|
| 1033 |
+
"epoch": 0.4170668823618624,
|
| 1034 |
+
"grad_norm": 4.2207112312316895,
|
| 1035 |
+
"learning_rate": 2.915573951783627e-05,
|
| 1036 |
+
"loss": 4.3184,
|
| 1037 |
+
"step": 13200
|
| 1038 |
+
},
|
| 1039 |
+
{
|
| 1040 |
+
"epoch": 0.4202264799555129,
|
| 1041 |
+
"grad_norm": 4.571977138519287,
|
| 1042 |
+
"learning_rate": 2.8999336471926447e-05,
|
| 1043 |
+
"loss": 4.2802,
|
| 1044 |
+
"step": 13300
|
| 1045 |
+
},
|
| 1046 |
+
{
|
| 1047 |
+
"epoch": 0.42338607754916335,
|
| 1048 |
+
"grad_norm": 4.995270252227783,
|
| 1049 |
+
"learning_rate": 2.884135359727006e-05,
|
| 1050 |
+
"loss": 4.3286,
|
| 1051 |
+
"step": 13400
|
| 1052 |
+
},
|
| 1053 |
+
{
|
| 1054 |
+
"epoch": 0.42654567514281383,
|
| 1055 |
+
"grad_norm": 3.6203622817993164,
|
| 1056 |
+
"learning_rate": 2.8683370722613672e-05,
|
| 1057 |
+
"loss": 4.3566,
|
| 1058 |
+
"step": 13500
|
| 1059 |
+
},
|
| 1060 |
+
{
|
| 1061 |
+
"epoch": 0.4297052727364643,
|
| 1062 |
+
"grad_norm": 3.7246546745300293,
|
| 1063 |
+
"learning_rate": 2.852538784795728e-05,
|
| 1064 |
+
"loss": 4.3191,
|
| 1065 |
+
"step": 13600
|
| 1066 |
+
},
|
| 1067 |
+
{
|
| 1068 |
+
"epoch": 0.4328648703301148,
|
| 1069 |
+
"grad_norm": 4.118934631347656,
|
| 1070 |
+
"learning_rate": 2.8367404973300894e-05,
|
| 1071 |
+
"loss": 4.3338,
|
| 1072 |
+
"step": 13700
|
| 1073 |
+
},
|
| 1074 |
+
{
|
| 1075 |
+
"epoch": 0.4360244679237652,
|
| 1076 |
+
"grad_norm": 4.569579601287842,
|
| 1077 |
+
"learning_rate": 2.8209422098644506e-05,
|
| 1078 |
+
"loss": 4.3436,
|
| 1079 |
+
"step": 13800
|
| 1080 |
+
},
|
| 1081 |
+
{
|
| 1082 |
+
"epoch": 0.4391840655174157,
|
| 1083 |
+
"grad_norm": 4.280219554901123,
|
| 1084 |
+
"learning_rate": 2.805143922398812e-05,
|
| 1085 |
+
"loss": 4.2796,
|
| 1086 |
+
"step": 13900
|
| 1087 |
+
},
|
| 1088 |
+
{
|
| 1089 |
+
"epoch": 0.44234366311106615,
|
| 1090 |
+
"grad_norm": 4.866397857666016,
|
| 1091 |
+
"learning_rate": 2.7893456349331732e-05,
|
| 1092 |
+
"loss": 4.314,
|
| 1093 |
+
"step": 14000
|
| 1094 |
+
},
|
| 1095 |
+
{
|
| 1096 |
+
"epoch": 0.44234366311106615,
|
| 1097 |
+
"eval_loss": 4.468830585479736,
|
| 1098 |
+
"eval_runtime": 707.6161,
|
| 1099 |
+
"eval_samples_per_second": 124.243,
|
| 1100 |
+
"eval_steps_per_second": 15.531,
|
| 1101 |
+
"step": 14000
|
| 1102 |
+
},
|
| 1103 |
+
{
|
| 1104 |
+
"epoch": 0.44550326070471663,
|
| 1105 |
+
"grad_norm": 4.688275337219238,
|
| 1106 |
+
"learning_rate": 2.7735473474675344e-05,
|
| 1107 |
+
"loss": 4.2977,
|
| 1108 |
+
"step": 14100
|
| 1109 |
+
},
|
| 1110 |
+
{
|
| 1111 |
+
"epoch": 0.4486628582983671,
|
| 1112 |
+
"grad_norm": 4.715358734130859,
|
| 1113 |
+
"learning_rate": 2.7577490600018957e-05,
|
| 1114 |
+
"loss": 4.357,
|
| 1115 |
+
"step": 14200
|
| 1116 |
+
},
|
| 1117 |
+
{
|
| 1118 |
+
"epoch": 0.4518224558920176,
|
| 1119 |
+
"grad_norm": 4.193299293518066,
|
| 1120 |
+
"learning_rate": 2.741950772536257e-05,
|
| 1121 |
+
"loss": 4.3002,
|
| 1122 |
+
"step": 14300
|
| 1123 |
+
},
|
| 1124 |
+
{
|
| 1125 |
+
"epoch": 0.45498205348566806,
|
| 1126 |
+
"grad_norm": 5.159686088562012,
|
| 1127 |
+
"learning_rate": 2.7261524850706182e-05,
|
| 1128 |
+
"loss": 4.2935,
|
| 1129 |
+
"step": 14400
|
| 1130 |
+
},
|
| 1131 |
+
{
|
| 1132 |
+
"epoch": 0.45814165107931853,
|
| 1133 |
+
"grad_norm": 4.735175609588623,
|
| 1134 |
+
"learning_rate": 2.7103541976049795e-05,
|
| 1135 |
+
"loss": 4.3297,
|
| 1136 |
+
"step": 14500
|
| 1137 |
+
},
|
| 1138 |
+
{
|
| 1139 |
+
"epoch": 0.461301248672969,
|
| 1140 |
+
"grad_norm": 4.604960918426514,
|
| 1141 |
+
"learning_rate": 2.6945559101393408e-05,
|
| 1142 |
+
"loss": 4.308,
|
| 1143 |
+
"step": 14600
|
| 1144 |
+
},
|
| 1145 |
+
{
|
| 1146 |
+
"epoch": 0.4644608462666195,
|
| 1147 |
+
"grad_norm": 4.662102222442627,
|
| 1148 |
+
"learning_rate": 2.678757622673702e-05,
|
| 1149 |
+
"loss": 4.2911,
|
| 1150 |
+
"step": 14700
|
| 1151 |
+
},
|
| 1152 |
+
{
|
| 1153 |
+
"epoch": 0.46762044386026996,
|
| 1154 |
+
"grad_norm": 5.6145477294921875,
|
| 1155 |
+
"learning_rate": 2.6629593352080633e-05,
|
| 1156 |
+
"loss": 4.283,
|
| 1157 |
+
"step": 14800
|
| 1158 |
+
},
|
| 1159 |
+
{
|
| 1160 |
+
"epoch": 0.47078004145392044,
|
| 1161 |
+
"grad_norm": 5.1714606285095215,
|
| 1162 |
+
"learning_rate": 2.6471610477424246e-05,
|
| 1163 |
+
"loss": 4.3117,
|
| 1164 |
+
"step": 14900
|
| 1165 |
+
},
|
| 1166 |
+
{
|
| 1167 |
+
"epoch": 0.4739396390475709,
|
| 1168 |
+
"grad_norm": 4.684764385223389,
|
| 1169 |
+
"learning_rate": 2.6313627602767858e-05,
|
| 1170 |
+
"loss": 4.3159,
|
| 1171 |
+
"step": 15000
|
| 1172 |
+
},
|
| 1173 |
+
{
|
| 1174 |
+
"epoch": 0.4739396390475709,
|
| 1175 |
+
"eval_loss": 4.460094451904297,
|
| 1176 |
+
"eval_runtime": 708.6547,
|
| 1177 |
+
"eval_samples_per_second": 124.06,
|
| 1178 |
+
"eval_steps_per_second": 15.508,
|
| 1179 |
+
"step": 15000
|
| 1180 |
+
},
|
| 1181 |
+
{
|
| 1182 |
+
"epoch": 0.4770992366412214,
|
| 1183 |
+
"grad_norm": 4.812125205993652,
|
| 1184 |
+
"learning_rate": 2.6155644728111474e-05,
|
| 1185 |
+
"loss": 4.297,
|
| 1186 |
+
"step": 15100
|
| 1187 |
+
},
|
| 1188 |
+
{
|
| 1189 |
+
"epoch": 0.48025883423487187,
|
| 1190 |
+
"grad_norm": 4.402428150177002,
|
| 1191 |
+
"learning_rate": 2.5997661853455087e-05,
|
| 1192 |
+
"loss": 4.2707,
|
| 1193 |
+
"step": 15200
|
| 1194 |
+
},
|
| 1195 |
+
{
|
| 1196 |
+
"epoch": 0.48341843182852234,
|
| 1197 |
+
"grad_norm": 3.6931350231170654,
|
| 1198 |
+
"learning_rate": 2.5841258807545265e-05,
|
| 1199 |
+
"loss": 4.2843,
|
| 1200 |
+
"step": 15300
|
| 1201 |
+
},
|
| 1202 |
+
{
|
| 1203 |
+
"epoch": 0.4865780294221728,
|
| 1204 |
+
"grad_norm": 4.092090129852295,
|
| 1205 |
+
"learning_rate": 2.5683275932888878e-05,
|
| 1206 |
+
"loss": 4.2672,
|
| 1207 |
+
"step": 15400
|
| 1208 |
+
},
|
| 1209 |
+
{
|
| 1210 |
+
"epoch": 0.48973762701582324,
|
| 1211 |
+
"grad_norm": 4.624776363372803,
|
| 1212 |
+
"learning_rate": 2.552529305823249e-05,
|
| 1213 |
+
"loss": 4.3225,
|
| 1214 |
+
"step": 15500
|
| 1215 |
+
},
|
| 1216 |
+
{
|
| 1217 |
+
"epoch": 0.4928972246094737,
|
| 1218 |
+
"grad_norm": 4.362007141113281,
|
| 1219 |
+
"learning_rate": 2.5367310183576103e-05,
|
| 1220 |
+
"loss": 4.3164,
|
| 1221 |
+
"step": 15600
|
| 1222 |
+
},
|
| 1223 |
+
{
|
| 1224 |
+
"epoch": 0.4960568222031242,
|
| 1225 |
+
"grad_norm": 3.67773175239563,
|
| 1226 |
+
"learning_rate": 2.5209327308919716e-05,
|
| 1227 |
+
"loss": 4.2915,
|
| 1228 |
+
"step": 15700
|
| 1229 |
+
},
|
| 1230 |
+
{
|
| 1231 |
+
"epoch": 0.49921641979677467,
|
| 1232 |
+
"grad_norm": 3.7847766876220703,
|
| 1233 |
+
"learning_rate": 2.505134443426333e-05,
|
| 1234 |
+
"loss": 4.2696,
|
| 1235 |
+
"step": 15800
|
| 1236 |
+
},
|
| 1237 |
+
{
|
| 1238 |
+
"epoch": 0.5023760173904251,
|
| 1239 |
+
"grad_norm": 4.13474178314209,
|
| 1240 |
+
"learning_rate": 2.489336155960694e-05,
|
| 1241 |
+
"loss": 4.3031,
|
| 1242 |
+
"step": 15900
|
| 1243 |
+
},
|
| 1244 |
+
{
|
| 1245 |
+
"epoch": 0.5055356149840756,
|
| 1246 |
+
"grad_norm": 4.143453598022461,
|
| 1247 |
+
"learning_rate": 2.4735378684950554e-05,
|
| 1248 |
+
"loss": 4.3384,
|
| 1249 |
+
"step": 16000
|
| 1250 |
+
},
|
| 1251 |
+
{
|
| 1252 |
+
"epoch": 0.5055356149840756,
|
| 1253 |
+
"eval_loss": 4.452681541442871,
|
| 1254 |
+
"eval_runtime": 706.2531,
|
| 1255 |
+
"eval_samples_per_second": 124.482,
|
| 1256 |
+
"eval_steps_per_second": 15.561,
|
| 1257 |
+
"step": 16000
|
| 1258 |
+
},
|
| 1259 |
+
{
|
| 1260 |
+
"epoch": 0.5086952125777261,
|
| 1261 |
+
"grad_norm": 4.447348594665527,
|
| 1262 |
+
"learning_rate": 2.4577395810294166e-05,
|
| 1263 |
+
"loss": 4.2958,
|
| 1264 |
+
"step": 16100
|
| 1265 |
+
},
|
| 1266 |
+
{
|
| 1267 |
+
"epoch": 0.5118548101713766,
|
| 1268 |
+
"grad_norm": 4.753751754760742,
|
| 1269 |
+
"learning_rate": 2.441941293563778e-05,
|
| 1270 |
+
"loss": 4.3129,
|
| 1271 |
+
"step": 16200
|
| 1272 |
+
},
|
| 1273 |
+
{
|
| 1274 |
+
"epoch": 0.515014407765027,
|
| 1275 |
+
"grad_norm": 3.83662748336792,
|
| 1276 |
+
"learning_rate": 2.426143006098139e-05,
|
| 1277 |
+
"loss": 4.3168,
|
| 1278 |
+
"step": 16300
|
| 1279 |
+
},
|
| 1280 |
+
{
|
| 1281 |
+
"epoch": 0.5181740053586775,
|
| 1282 |
+
"grad_norm": 3.786252737045288,
|
| 1283 |
+
"learning_rate": 2.4103447186325004e-05,
|
| 1284 |
+
"loss": 4.2844,
|
| 1285 |
+
"step": 16400
|
| 1286 |
+
},
|
| 1287 |
+
{
|
| 1288 |
+
"epoch": 0.521333602952328,
|
| 1289 |
+
"grad_norm": 3.9351420402526855,
|
| 1290 |
+
"learning_rate": 2.3945464311668617e-05,
|
| 1291 |
+
"loss": 4.2816,
|
| 1292 |
+
"step": 16500
|
| 1293 |
+
},
|
| 1294 |
+
{
|
| 1295 |
+
"epoch": 0.5244932005459785,
|
| 1296 |
+
"grad_norm": 4.529867172241211,
|
| 1297 |
+
"learning_rate": 2.378748143701223e-05,
|
| 1298 |
+
"loss": 4.302,
|
| 1299 |
+
"step": 16600
|
| 1300 |
+
},
|
| 1301 |
+
{
|
| 1302 |
+
"epoch": 0.527652798139629,
|
| 1303 |
+
"grad_norm": 4.681921005249023,
|
| 1304 |
+
"learning_rate": 2.3629498562355842e-05,
|
| 1305 |
+
"loss": 4.2904,
|
| 1306 |
+
"step": 16700
|
| 1307 |
+
},
|
| 1308 |
+
{
|
| 1309 |
+
"epoch": 0.5308123957332794,
|
| 1310 |
+
"grad_norm": 3.3952362537384033,
|
| 1311 |
+
"learning_rate": 2.3471515687699455e-05,
|
| 1312 |
+
"loss": 4.2876,
|
| 1313 |
+
"step": 16800
|
| 1314 |
+
},
|
| 1315 |
+
{
|
| 1316 |
+
"epoch": 0.5339719933269299,
|
| 1317 |
+
"grad_norm": 4.524105548858643,
|
| 1318 |
+
"learning_rate": 2.3313532813043067e-05,
|
| 1319 |
+
"loss": 4.262,
|
| 1320 |
+
"step": 16900
|
| 1321 |
+
},
|
| 1322 |
+
{
|
| 1323 |
+
"epoch": 0.5371315909205804,
|
| 1324 |
+
"grad_norm": 4.248294830322266,
|
| 1325 |
+
"learning_rate": 2.315554993838668e-05,
|
| 1326 |
+
"loss": 4.2975,
|
| 1327 |
+
"step": 17000
|
| 1328 |
+
},
|
| 1329 |
+
{
|
| 1330 |
+
"epoch": 0.5371315909205804,
|
| 1331 |
+
"eval_loss": 4.445527076721191,
|
| 1332 |
+
"eval_runtime": 705.3764,
|
| 1333 |
+
"eval_samples_per_second": 124.637,
|
| 1334 |
+
"eval_steps_per_second": 15.58,
|
| 1335 |
+
"step": 17000
|
| 1336 |
+
},
|
| 1337 |
+
{
|
| 1338 |
+
"epoch": 0.5402911885142309,
|
| 1339 |
+
"grad_norm": 4.230813026428223,
|
| 1340 |
+
"learning_rate": 2.2997567063730293e-05,
|
| 1341 |
+
"loss": 4.2806,
|
| 1342 |
+
"step": 17100
|
| 1343 |
+
},
|
| 1344 |
+
{
|
| 1345 |
+
"epoch": 0.5434507861078813,
|
| 1346 |
+
"grad_norm": 3.7945194244384766,
|
| 1347 |
+
"learning_rate": 2.2839584189073905e-05,
|
| 1348 |
+
"loss": 4.2942,
|
| 1349 |
+
"step": 17200
|
| 1350 |
+
},
|
| 1351 |
+
{
|
| 1352 |
+
"epoch": 0.5466103837015318,
|
| 1353 |
+
"grad_norm": 4.855790138244629,
|
| 1354 |
+
"learning_rate": 2.2681601314417518e-05,
|
| 1355 |
+
"loss": 4.2818,
|
| 1356 |
+
"step": 17300
|
| 1357 |
+
},
|
| 1358 |
+
{
|
| 1359 |
+
"epoch": 0.5497699812951823,
|
| 1360 |
+
"grad_norm": 4.7962236404418945,
|
| 1361 |
+
"learning_rate": 2.252361843976113e-05,
|
| 1362 |
+
"loss": 4.2804,
|
| 1363 |
+
"step": 17400
|
| 1364 |
+
},
|
| 1365 |
+
{
|
| 1366 |
+
"epoch": 0.5529295788888328,
|
| 1367 |
+
"grad_norm": 4.337451934814453,
|
| 1368 |
+
"learning_rate": 2.2365635565104747e-05,
|
| 1369 |
+
"loss": 4.3063,
|
| 1370 |
+
"step": 17500
|
| 1371 |
+
},
|
| 1372 |
+
{
|
| 1373 |
+
"epoch": 0.5560891764824832,
|
| 1374 |
+
"grad_norm": 4.422916889190674,
|
| 1375 |
+
"learning_rate": 2.220765269044836e-05,
|
| 1376 |
+
"loss": 4.2942,
|
| 1377 |
+
"step": 17600
|
| 1378 |
+
},
|
| 1379 |
+
{
|
| 1380 |
+
"epoch": 0.5592487740761337,
|
| 1381 |
+
"grad_norm": 4.27534294128418,
|
| 1382 |
+
"learning_rate": 2.204966981579197e-05,
|
| 1383 |
+
"loss": 4.2828,
|
| 1384 |
+
"step": 17700
|
| 1385 |
+
},
|
| 1386 |
+
{
|
| 1387 |
+
"epoch": 0.5624083716697842,
|
| 1388 |
+
"grad_norm": 4.3357648849487305,
|
| 1389 |
+
"learning_rate": 2.1893266769882147e-05,
|
| 1390 |
+
"loss": 4.2766,
|
| 1391 |
+
"step": 17800
|
| 1392 |
+
},
|
| 1393 |
+
{
|
| 1394 |
+
"epoch": 0.5655679692634346,
|
| 1395 |
+
"grad_norm": 3.859369993209839,
|
| 1396 |
+
"learning_rate": 2.173528389522576e-05,
|
| 1397 |
+
"loss": 4.2979,
|
| 1398 |
+
"step": 17900
|
| 1399 |
+
},
|
| 1400 |
+
{
|
| 1401 |
+
"epoch": 0.568727566857085,
|
| 1402 |
+
"grad_norm": 5.381210803985596,
|
| 1403 |
+
"learning_rate": 2.1577301020569372e-05,
|
| 1404 |
+
"loss": 4.3169,
|
| 1405 |
+
"step": 18000
|
| 1406 |
+
},
|
| 1407 |
+
{
|
| 1408 |
+
"epoch": 0.568727566857085,
|
| 1409 |
+
"eval_loss": 4.439344882965088,
|
| 1410 |
+
"eval_runtime": 708.2829,
|
| 1411 |
+
"eval_samples_per_second": 124.126,
|
| 1412 |
+
"eval_steps_per_second": 15.516,
|
| 1413 |
+
"step": 18000
|
| 1414 |
+
},
|
| 1415 |
+
{
|
| 1416 |
+
"epoch": 0.5718871644507355,
|
| 1417 |
+
"grad_norm": 3.3209519386291504,
|
| 1418 |
+
"learning_rate": 2.1419318145912985e-05,
|
| 1419 |
+
"loss": 4.2588,
|
| 1420 |
+
"step": 18100
|
| 1421 |
+
},
|
| 1422 |
+
{
|
| 1423 |
+
"epoch": 0.575046762044386,
|
| 1424 |
+
"grad_norm": 3.9373040199279785,
|
| 1425 |
+
"learning_rate": 2.1261335271256597e-05,
|
| 1426 |
+
"loss": 4.2895,
|
| 1427 |
+
"step": 18200
|
| 1428 |
+
},
|
| 1429 |
+
{
|
| 1430 |
+
"epoch": 0.5782063596380365,
|
| 1431 |
+
"grad_norm": 3.453148126602173,
|
| 1432 |
+
"learning_rate": 2.110335239660021e-05,
|
| 1433 |
+
"loss": 4.276,
|
| 1434 |
+
"step": 18300
|
| 1435 |
+
},
|
| 1436 |
+
{
|
| 1437 |
+
"epoch": 0.5813659572316869,
|
| 1438 |
+
"grad_norm": 3.7350664138793945,
|
| 1439 |
+
"learning_rate": 2.0945369521943823e-05,
|
| 1440 |
+
"loss": 4.3268,
|
| 1441 |
+
"step": 18400
|
| 1442 |
+
},
|
| 1443 |
+
{
|
| 1444 |
+
"epoch": 0.5845255548253374,
|
| 1445 |
+
"grad_norm": 4.277946949005127,
|
| 1446 |
+
"learning_rate": 2.0787386647287435e-05,
|
| 1447 |
+
"loss": 4.2573,
|
| 1448 |
+
"step": 18500
|
| 1449 |
+
},
|
| 1450 |
+
{
|
| 1451 |
+
"epoch": 0.5876851524189879,
|
| 1452 |
+
"grad_norm": 4.000477313995361,
|
| 1453 |
+
"learning_rate": 2.0629403772631048e-05,
|
| 1454 |
+
"loss": 4.2568,
|
| 1455 |
+
"step": 18600
|
| 1456 |
+
},
|
| 1457 |
+
{
|
| 1458 |
+
"epoch": 0.5908447500126384,
|
| 1459 |
+
"grad_norm": 3.6891930103302,
|
| 1460 |
+
"learning_rate": 2.047142089797466e-05,
|
| 1461 |
+
"loss": 4.2784,
|
| 1462 |
+
"step": 18700
|
| 1463 |
+
},
|
| 1464 |
+
{
|
| 1465 |
+
"epoch": 0.5940043476062888,
|
| 1466 |
+
"grad_norm": 4.800564765930176,
|
| 1467 |
+
"learning_rate": 2.0313438023318273e-05,
|
| 1468 |
+
"loss": 4.3016,
|
| 1469 |
+
"step": 18800
|
| 1470 |
+
},
|
| 1471 |
+
{
|
| 1472 |
+
"epoch": 0.5971639451999393,
|
| 1473 |
+
"grad_norm": 5.003584384918213,
|
| 1474 |
+
"learning_rate": 2.0155455148661886e-05,
|
| 1475 |
+
"loss": 4.2818,
|
| 1476 |
+
"step": 18900
|
| 1477 |
+
},
|
| 1478 |
+
{
|
| 1479 |
+
"epoch": 0.6003235427935898,
|
| 1480 |
+
"grad_norm": 4.17577600479126,
|
| 1481 |
+
"learning_rate": 1.99974722740055e-05,
|
| 1482 |
+
"loss": 4.2794,
|
| 1483 |
+
"step": 19000
|
| 1484 |
+
},
|
| 1485 |
+
{
|
| 1486 |
+
"epoch": 0.6003235427935898,
|
| 1487 |
+
"eval_loss": 4.433346748352051,
|
| 1488 |
+
"eval_runtime": 712.1597,
|
| 1489 |
+
"eval_samples_per_second": 123.45,
|
| 1490 |
+
"eval_steps_per_second": 15.432,
|
| 1491 |
+
"step": 19000
|
| 1492 |
+
},
|
| 1493 |
+
{
|
| 1494 |
+
"epoch": 0.6034831403872403,
|
| 1495 |
+
"grad_norm": 4.12880277633667,
|
| 1496 |
+
"learning_rate": 1.983948939934911e-05,
|
| 1497 |
+
"loss": 4.2539,
|
| 1498 |
+
"step": 19100
|
| 1499 |
+
},
|
| 1500 |
+
{
|
| 1501 |
+
"epoch": 0.6066427379808907,
|
| 1502 |
+
"grad_norm": 4.599102020263672,
|
| 1503 |
+
"learning_rate": 1.9681506524692724e-05,
|
| 1504 |
+
"loss": 4.2842,
|
| 1505 |
+
"step": 19200
|
| 1506 |
+
},
|
| 1507 |
+
{
|
| 1508 |
+
"epoch": 0.6098023355745412,
|
| 1509 |
+
"grad_norm": 4.2362380027771,
|
| 1510 |
+
"learning_rate": 1.9523523650036336e-05,
|
| 1511 |
+
"loss": 4.2803,
|
| 1512 |
+
"step": 19300
|
| 1513 |
+
},
|
| 1514 |
+
{
|
| 1515 |
+
"epoch": 0.6129619331681917,
|
| 1516 |
+
"grad_norm": 4.119621753692627,
|
| 1517 |
+
"learning_rate": 1.936554077537995e-05,
|
| 1518 |
+
"loss": 4.2413,
|
| 1519 |
+
"step": 19400
|
| 1520 |
+
},
|
| 1521 |
+
{
|
| 1522 |
+
"epoch": 0.6161215307618422,
|
| 1523 |
+
"grad_norm": 4.374582290649414,
|
| 1524 |
+
"learning_rate": 1.9207557900723565e-05,
|
| 1525 |
+
"loss": 4.3056,
|
| 1526 |
+
"step": 19500
|
| 1527 |
+
},
|
| 1528 |
+
{
|
| 1529 |
+
"epoch": 0.6192811283554926,
|
| 1530 |
+
"grad_norm": 3.5877952575683594,
|
| 1531 |
+
"learning_rate": 1.9049575026067178e-05,
|
| 1532 |
+
"loss": 4.2978,
|
| 1533 |
+
"step": 19600
|
| 1534 |
+
},
|
| 1535 |
+
{
|
| 1536 |
+
"epoch": 0.6224407259491431,
|
| 1537 |
+
"grad_norm": 4.062015056610107,
|
| 1538 |
+
"learning_rate": 1.889159215141079e-05,
|
| 1539 |
+
"loss": 4.2525,
|
| 1540 |
+
"step": 19700
|
| 1541 |
+
},
|
| 1542 |
+
{
|
| 1543 |
+
"epoch": 0.6256003235427936,
|
| 1544 |
+
"grad_norm": 4.872693061828613,
|
| 1545 |
+
"learning_rate": 1.8733609276754403e-05,
|
| 1546 |
+
"loss": 4.3039,
|
| 1547 |
+
"step": 19800
|
| 1548 |
+
},
|
| 1549 |
+
{
|
| 1550 |
+
"epoch": 0.6287599211364441,
|
| 1551 |
+
"grad_norm": 3.899491310119629,
|
| 1552 |
+
"learning_rate": 1.8575626402098016e-05,
|
| 1553 |
+
"loss": 4.3016,
|
| 1554 |
+
"step": 19900
|
| 1555 |
+
},
|
| 1556 |
+
{
|
| 1557 |
+
"epoch": 0.6319195187300946,
|
| 1558 |
+
"grad_norm": 3.927971601486206,
|
| 1559 |
+
"learning_rate": 1.8417643527441628e-05,
|
| 1560 |
+
"loss": 4.2883,
|
| 1561 |
+
"step": 20000
|
| 1562 |
+
},
|
| 1563 |
+
{
|
| 1564 |
+
"epoch": 0.6319195187300946,
|
| 1565 |
+
"eval_loss": 4.4285101890563965,
|
| 1566 |
+
"eval_runtime": 708.5523,
|
| 1567 |
+
"eval_samples_per_second": 124.078,
|
| 1568 |
+
"eval_steps_per_second": 15.51,
|
| 1569 |
+
"step": 20000
|
| 1570 |
+
},
|
| 1571 |
+
{
|
| 1572 |
+
"epoch": 0.635079116323745,
|
| 1573 |
+
"grad_norm": 3.7342617511749268,
|
| 1574 |
+
"learning_rate": 1.8259660652785237e-05,
|
| 1575 |
+
"loss": 4.2633,
|
| 1576 |
+
"step": 20100
|
| 1577 |
+
},
|
| 1578 |
+
{
|
| 1579 |
+
"epoch": 0.6382387139173955,
|
| 1580 |
+
"grad_norm": 5.151529788970947,
|
| 1581 |
+
"learning_rate": 1.810167777812885e-05,
|
| 1582 |
+
"loss": 4.2858,
|
| 1583 |
+
"step": 20200
|
| 1584 |
+
},
|
| 1585 |
+
{
|
| 1586 |
+
"epoch": 0.641398311511046,
|
| 1587 |
+
"grad_norm": 6.981259822845459,
|
| 1588 |
+
"learning_rate": 1.7943694903472463e-05,
|
| 1589 |
+
"loss": 4.2812,
|
| 1590 |
+
"step": 20300
|
| 1591 |
+
},
|
| 1592 |
+
{
|
| 1593 |
+
"epoch": 0.6445579091046965,
|
| 1594 |
+
"grad_norm": 4.146710395812988,
|
| 1595 |
+
"learning_rate": 1.7785712028816075e-05,
|
| 1596 |
+
"loss": 4.295,
|
| 1597 |
+
"step": 20400
|
| 1598 |
+
},
|
| 1599 |
+
{
|
| 1600 |
+
"epoch": 0.6477175066983469,
|
| 1601 |
+
"grad_norm": 3.6891162395477295,
|
| 1602 |
+
"learning_rate": 1.7629308982906254e-05,
|
| 1603 |
+
"loss": 4.268,
|
| 1604 |
+
"step": 20500
|
| 1605 |
+
},
|
| 1606 |
+
{
|
| 1607 |
+
"epoch": 0.6508771042919974,
|
| 1608 |
+
"grad_norm": 4.074377536773682,
|
| 1609 |
+
"learning_rate": 1.7471326108249866e-05,
|
| 1610 |
+
"loss": 4.2796,
|
| 1611 |
+
"step": 20600
|
| 1612 |
+
},
|
| 1613 |
+
{
|
| 1614 |
+
"epoch": 0.6540367018856479,
|
| 1615 |
+
"grad_norm": 3.6414151191711426,
|
| 1616 |
+
"learning_rate": 1.731334323359348e-05,
|
| 1617 |
+
"loss": 4.2815,
|
| 1618 |
+
"step": 20700
|
| 1619 |
+
},
|
| 1620 |
+
{
|
| 1621 |
+
"epoch": 0.6571962994792984,
|
| 1622 |
+
"grad_norm": 4.212449073791504,
|
| 1623 |
+
"learning_rate": 1.7156940187683657e-05,
|
| 1624 |
+
"loss": 4.2843,
|
| 1625 |
+
"step": 20800
|
| 1626 |
+
},
|
| 1627 |
+
{
|
| 1628 |
+
"epoch": 0.6603558970729488,
|
| 1629 |
+
"grad_norm": 4.951251983642578,
|
| 1630 |
+
"learning_rate": 1.699895731302727e-05,
|
| 1631 |
+
"loss": 4.271,
|
| 1632 |
+
"step": 20900
|
| 1633 |
+
},
|
| 1634 |
+
{
|
| 1635 |
+
"epoch": 0.6635154946665993,
|
| 1636 |
+
"grad_norm": 5.257840633392334,
|
| 1637 |
+
"learning_rate": 1.6840974438370883e-05,
|
| 1638 |
+
"loss": 4.2744,
|
| 1639 |
+
"step": 21000
|
| 1640 |
+
},
|
| 1641 |
+
{
|
| 1642 |
+
"epoch": 0.6635154946665993,
|
| 1643 |
+
"eval_loss": 4.42351770401001,
|
| 1644 |
+
"eval_runtime": 703.9002,
|
| 1645 |
+
"eval_samples_per_second": 124.898,
|
| 1646 |
+
"eval_steps_per_second": 15.613,
|
| 1647 |
+
"step": 21000
|
| 1648 |
+
},
|
| 1649 |
+
{
|
| 1650 |
+
"epoch": 0.6666750922602497,
|
| 1651 |
+
"grad_norm": 4.769164562225342,
|
| 1652 |
+
"learning_rate": 1.6682991563714495e-05,
|
| 1653 |
+
"loss": 4.2332,
|
| 1654 |
+
"step": 21100
|
| 1655 |
+
},
|
| 1656 |
+
{
|
| 1657 |
+
"epoch": 0.6698346898539002,
|
| 1658 |
+
"grad_norm": 3.9750113487243652,
|
| 1659 |
+
"learning_rate": 1.6525008689058108e-05,
|
| 1660 |
+
"loss": 4.256,
|
| 1661 |
+
"step": 21200
|
| 1662 |
+
},
|
| 1663 |
+
{
|
| 1664 |
+
"epoch": 0.6729942874475506,
|
| 1665 |
+
"grad_norm": 4.802659511566162,
|
| 1666 |
+
"learning_rate": 1.636702581440172e-05,
|
| 1667 |
+
"loss": 4.2465,
|
| 1668 |
+
"step": 21300
|
| 1669 |
+
},
|
| 1670 |
+
{
|
| 1671 |
+
"epoch": 0.6761538850412011,
|
| 1672 |
+
"grad_norm": 4.48084831237793,
|
| 1673 |
+
"learning_rate": 1.6209042939745333e-05,
|
| 1674 |
+
"loss": 4.2762,
|
| 1675 |
+
"step": 21400
|
| 1676 |
+
},
|
| 1677 |
+
{
|
| 1678 |
+
"epoch": 0.6793134826348516,
|
| 1679 |
+
"grad_norm": 5.000456809997559,
|
| 1680 |
+
"learning_rate": 1.6051060065088946e-05,
|
| 1681 |
+
"loss": 4.2937,
|
| 1682 |
+
"step": 21500
|
| 1683 |
+
},
|
| 1684 |
+
{
|
| 1685 |
+
"epoch": 0.6824730802285021,
|
| 1686 |
+
"grad_norm": 4.686836242675781,
|
| 1687 |
+
"learning_rate": 1.5893077190432555e-05,
|
| 1688 |
+
"loss": 4.2684,
|
| 1689 |
+
"step": 21600
|
| 1690 |
+
},
|
| 1691 |
+
{
|
| 1692 |
+
"epoch": 0.6856326778221525,
|
| 1693 |
+
"grad_norm": 3.528893232345581,
|
| 1694 |
+
"learning_rate": 1.5735094315776168e-05,
|
| 1695 |
+
"loss": 4.2491,
|
| 1696 |
+
"step": 21700
|
| 1697 |
+
},
|
| 1698 |
+
{
|
| 1699 |
+
"epoch": 0.688792275415803,
|
| 1700 |
+
"grad_norm": 4.837214946746826,
|
| 1701 |
+
"learning_rate": 1.5577111441119784e-05,
|
| 1702 |
+
"loss": 4.2601,
|
| 1703 |
+
"step": 21800
|
| 1704 |
+
},
|
| 1705 |
+
{
|
| 1706 |
+
"epoch": 0.6919518730094535,
|
| 1707 |
+
"grad_norm": 4.445570468902588,
|
| 1708 |
+
"learning_rate": 1.5419128566463396e-05,
|
| 1709 |
+
"loss": 4.2533,
|
| 1710 |
+
"step": 21900
|
| 1711 |
+
},
|
| 1712 |
+
{
|
| 1713 |
+
"epoch": 0.695111470603104,
|
| 1714 |
+
"grad_norm": 4.073211193084717,
|
| 1715 |
+
"learning_rate": 1.526114569180701e-05,
|
| 1716 |
+
"loss": 4.2732,
|
| 1717 |
+
"step": 22000
|
| 1718 |
+
},
|
| 1719 |
+
{
|
| 1720 |
+
"epoch": 0.695111470603104,
|
| 1721 |
+
"eval_loss": 4.418666839599609,
|
| 1722 |
+
"eval_runtime": 706.2044,
|
| 1723 |
+
"eval_samples_per_second": 124.491,
|
| 1724 |
+
"eval_steps_per_second": 15.562,
|
| 1725 |
+
"step": 22000
|
| 1726 |
+
},
|
| 1727 |
+
{
|
| 1728 |
+
"epoch": 0.6982710681967544,
|
| 1729 |
+
"grad_norm": 4.091275691986084,
|
| 1730 |
+
"learning_rate": 1.510316281715062e-05,
|
| 1731 |
+
"loss": 4.2226,
|
| 1732 |
+
"step": 22100
|
| 1733 |
+
},
|
| 1734 |
+
{
|
| 1735 |
+
"epoch": 0.7014306657904049,
|
| 1736 |
+
"grad_norm": 4.8102192878723145,
|
| 1737 |
+
"learning_rate": 1.4945179942494233e-05,
|
| 1738 |
+
"loss": 4.2794,
|
| 1739 |
+
"step": 22200
|
| 1740 |
+
},
|
| 1741 |
+
{
|
| 1742 |
+
"epoch": 0.7045902633840554,
|
| 1743 |
+
"grad_norm": 4.450103282928467,
|
| 1744 |
+
"learning_rate": 1.4787197067837847e-05,
|
| 1745 |
+
"loss": 4.2869,
|
| 1746 |
+
"step": 22300
|
| 1747 |
+
},
|
| 1748 |
+
{
|
| 1749 |
+
"epoch": 0.7077498609777059,
|
| 1750 |
+
"grad_norm": 4.858384609222412,
|
| 1751 |
+
"learning_rate": 1.462921419318146e-05,
|
| 1752 |
+
"loss": 4.2616,
|
| 1753 |
+
"step": 22400
|
| 1754 |
+
},
|
| 1755 |
+
{
|
| 1756 |
+
"epoch": 0.7109094585713563,
|
| 1757 |
+
"grad_norm": 4.810959339141846,
|
| 1758 |
+
"learning_rate": 1.4471231318525072e-05,
|
| 1759 |
+
"loss": 4.2589,
|
| 1760 |
+
"step": 22500
|
| 1761 |
+
},
|
| 1762 |
+
{
|
| 1763 |
+
"epoch": 0.7140690561650068,
|
| 1764 |
+
"grad_norm": 4.832188129425049,
|
| 1765 |
+
"learning_rate": 1.4313248443868685e-05,
|
| 1766 |
+
"loss": 4.2668,
|
| 1767 |
+
"step": 22600
|
| 1768 |
+
},
|
| 1769 |
+
{
|
| 1770 |
+
"epoch": 0.7172286537586573,
|
| 1771 |
+
"grad_norm": 4.8991475105285645,
|
| 1772 |
+
"learning_rate": 1.4155265569212298e-05,
|
| 1773 |
+
"loss": 4.2759,
|
| 1774 |
+
"step": 22700
|
| 1775 |
+
},
|
| 1776 |
+
{
|
| 1777 |
+
"epoch": 0.7203882513523078,
|
| 1778 |
+
"grad_norm": 5.653695106506348,
|
| 1779 |
+
"learning_rate": 1.399728269455591e-05,
|
| 1780 |
+
"loss": 4.2745,
|
| 1781 |
+
"step": 22800
|
| 1782 |
+
},
|
| 1783 |
+
{
|
| 1784 |
+
"epoch": 0.7235478489459582,
|
| 1785 |
+
"grad_norm": 4.475470542907715,
|
| 1786 |
+
"learning_rate": 1.3839299819899523e-05,
|
| 1787 |
+
"loss": 4.2785,
|
| 1788 |
+
"step": 22900
|
| 1789 |
+
},
|
| 1790 |
+
{
|
| 1791 |
+
"epoch": 0.7267074465396087,
|
| 1792 |
+
"grad_norm": 4.225826263427734,
|
| 1793 |
+
"learning_rate": 1.3681316945243135e-05,
|
| 1794 |
+
"loss": 4.257,
|
| 1795 |
+
"step": 23000
|
| 1796 |
+
},
|
| 1797 |
+
{
|
| 1798 |
+
"epoch": 0.7267074465396087,
|
| 1799 |
+
"eval_loss": 4.414594650268555,
|
| 1800 |
+
"eval_runtime": 709.538,
|
| 1801 |
+
"eval_samples_per_second": 123.906,
|
| 1802 |
+
"eval_steps_per_second": 15.489,
|
| 1803 |
+
"step": 23000
|
| 1804 |
+
},
|
| 1805 |
+
{
|
| 1806 |
+
"epoch": 0.7298670441332592,
|
| 1807 |
+
"grad_norm": 5.017648220062256,
|
| 1808 |
+
"learning_rate": 1.352333407058675e-05,
|
| 1809 |
+
"loss": 4.3036,
|
| 1810 |
+
"step": 23100
|
| 1811 |
+
},
|
| 1812 |
+
{
|
| 1813 |
+
"epoch": 0.7330266417269097,
|
| 1814 |
+
"grad_norm": 4.95635986328125,
|
| 1815 |
+
"learning_rate": 1.3365351195930362e-05,
|
| 1816 |
+
"loss": 4.295,
|
| 1817 |
+
"step": 23200
|
| 1818 |
+
},
|
| 1819 |
+
{
|
| 1820 |
+
"epoch": 0.7361862393205602,
|
| 1821 |
+
"grad_norm": 5.106438159942627,
|
| 1822 |
+
"learning_rate": 1.3207368321273975e-05,
|
| 1823 |
+
"loss": 4.2452,
|
| 1824 |
+
"step": 23300
|
| 1825 |
+
},
|
| 1826 |
+
{
|
| 1827 |
+
"epoch": 0.7393458369142106,
|
| 1828 |
+
"grad_norm": 5.10938024520874,
|
| 1829 |
+
"learning_rate": 1.3049385446617588e-05,
|
| 1830 |
+
"loss": 4.2532,
|
| 1831 |
+
"step": 23400
|
| 1832 |
+
},
|
| 1833 |
+
{
|
| 1834 |
+
"epoch": 0.7425054345078611,
|
| 1835 |
+
"grad_norm": 4.140163421630859,
|
| 1836 |
+
"learning_rate": 1.28914025719612e-05,
|
| 1837 |
+
"loss": 4.262,
|
| 1838 |
+
"step": 23500
|
| 1839 |
+
},
|
| 1840 |
+
{
|
| 1841 |
+
"epoch": 0.7456650321015116,
|
| 1842 |
+
"grad_norm": 4.146867275238037,
|
| 1843 |
+
"learning_rate": 1.2733419697304813e-05,
|
| 1844 |
+
"loss": 4.2381,
|
| 1845 |
+
"step": 23600
|
| 1846 |
+
},
|
| 1847 |
+
{
|
| 1848 |
+
"epoch": 0.7488246296951621,
|
| 1849 |
+
"grad_norm": 4.620384216308594,
|
| 1850 |
+
"learning_rate": 1.2575436822648426e-05,
|
| 1851 |
+
"loss": 4.2509,
|
| 1852 |
+
"step": 23700
|
| 1853 |
+
},
|
| 1854 |
+
{
|
| 1855 |
+
"epoch": 0.7519842272888125,
|
| 1856 |
+
"grad_norm": 4.155729293823242,
|
| 1857 |
+
"learning_rate": 1.2417453947992038e-05,
|
| 1858 |
+
"loss": 4.2816,
|
| 1859 |
+
"step": 23800
|
| 1860 |
+
},
|
| 1861 |
+
{
|
| 1862 |
+
"epoch": 0.755143824882463,
|
| 1863 |
+
"grad_norm": 6.052363395690918,
|
| 1864 |
+
"learning_rate": 1.2259471073335651e-05,
|
| 1865 |
+
"loss": 4.2619,
|
| 1866 |
+
"step": 23900
|
| 1867 |
+
},
|
| 1868 |
+
{
|
| 1869 |
+
"epoch": 0.7583034224761135,
|
| 1870 |
+
"grad_norm": 4.320205211639404,
|
| 1871 |
+
"learning_rate": 1.2101488198679265e-05,
|
| 1872 |
+
"loss": 4.2691,
|
| 1873 |
+
"step": 24000
|
| 1874 |
+
},
|
| 1875 |
+
{
|
| 1876 |
+
"epoch": 0.7583034224761135,
|
| 1877 |
+
"eval_loss": 4.410797595977783,
|
| 1878 |
+
"eval_runtime": 708.2786,
|
| 1879 |
+
"eval_samples_per_second": 124.126,
|
| 1880 |
+
"eval_steps_per_second": 15.516,
|
| 1881 |
+
"step": 24000
|
| 1882 |
+
},
|
| 1883 |
+
{
|
| 1884 |
+
"epoch": 0.761463020069764,
|
| 1885 |
+
"grad_norm": 4.142159461975098,
|
| 1886 |
+
"learning_rate": 1.1943505324022876e-05,
|
| 1887 |
+
"loss": 4.277,
|
| 1888 |
+
"step": 24100
|
| 1889 |
+
},
|
| 1890 |
+
{
|
| 1891 |
+
"epoch": 0.7646226176634144,
|
| 1892 |
+
"grad_norm": 4.465941905975342,
|
| 1893 |
+
"learning_rate": 1.1785522449366489e-05,
|
| 1894 |
+
"loss": 4.2629,
|
| 1895 |
+
"step": 24200
|
| 1896 |
+
},
|
| 1897 |
+
{
|
| 1898 |
+
"epoch": 0.7677822152570649,
|
| 1899 |
+
"grad_norm": 3.3799822330474854,
|
| 1900 |
+
"learning_rate": 1.1627539574710101e-05,
|
| 1901 |
+
"loss": 4.2882,
|
| 1902 |
+
"step": 24300
|
| 1903 |
+
},
|
| 1904 |
+
{
|
| 1905 |
+
"epoch": 0.7709418128507153,
|
| 1906 |
+
"grad_norm": 4.531765460968018,
|
| 1907 |
+
"learning_rate": 1.1469556700053714e-05,
|
| 1908 |
+
"loss": 4.2546,
|
| 1909 |
+
"step": 24400
|
| 1910 |
+
},
|
| 1911 |
+
{
|
| 1912 |
+
"epoch": 0.7741014104443658,
|
| 1913 |
+
"grad_norm": 4.775002956390381,
|
| 1914 |
+
"learning_rate": 1.1311573825397327e-05,
|
| 1915 |
+
"loss": 4.2376,
|
| 1916 |
+
"step": 24500
|
| 1917 |
+
},
|
| 1918 |
+
{
|
| 1919 |
+
"epoch": 0.7772610080380162,
|
| 1920 |
+
"grad_norm": 5.156617164611816,
|
| 1921 |
+
"learning_rate": 1.115359095074094e-05,
|
| 1922 |
+
"loss": 4.2341,
|
| 1923 |
+
"step": 24600
|
| 1924 |
+
},
|
| 1925 |
+
{
|
| 1926 |
+
"epoch": 0.7804206056316667,
|
| 1927 |
+
"grad_norm": 4.456455230712891,
|
| 1928 |
+
"learning_rate": 1.0995608076084554e-05,
|
| 1929 |
+
"loss": 4.2659,
|
| 1930 |
+
"step": 24700
|
| 1931 |
+
},
|
| 1932 |
+
{
|
| 1933 |
+
"epoch": 0.7835802032253172,
|
| 1934 |
+
"grad_norm": 4.179417133331299,
|
| 1935 |
+
"learning_rate": 1.083920503017473e-05,
|
| 1936 |
+
"loss": 4.2722,
|
| 1937 |
+
"step": 24800
|
| 1938 |
+
},
|
| 1939 |
+
{
|
| 1940 |
+
"epoch": 0.7867398008189677,
|
| 1941 |
+
"grad_norm": 4.3775224685668945,
|
| 1942 |
+
"learning_rate": 1.0681222155518341e-05,
|
| 1943 |
+
"loss": 4.278,
|
| 1944 |
+
"step": 24900
|
| 1945 |
+
},
|
| 1946 |
+
{
|
| 1947 |
+
"epoch": 0.7898993984126181,
|
| 1948 |
+
"grad_norm": 3.773751974105835,
|
| 1949 |
+
"learning_rate": 1.0523239280861954e-05,
|
| 1950 |
+
"loss": 4.2548,
|
| 1951 |
+
"step": 25000
|
| 1952 |
+
},
|
| 1953 |
+
{
|
| 1954 |
+
"epoch": 0.7898993984126181,
|
| 1955 |
+
"eval_loss": 4.407622814178467,
|
| 1956 |
+
"eval_runtime": 708.246,
|
| 1957 |
+
"eval_samples_per_second": 124.132,
|
| 1958 |
+
"eval_steps_per_second": 15.517,
|
| 1959 |
+
"step": 25000
|
| 1960 |
+
},
|
| 1961 |
+
{
|
| 1962 |
+
"epoch": 0.7930589960062686,
|
| 1963 |
+
"grad_norm": 3.457412004470825,
|
| 1964 |
+
"learning_rate": 1.0365256406205567e-05,
|
| 1965 |
+
"loss": 4.2549,
|
| 1966 |
+
"step": 25100
|
| 1967 |
+
},
|
| 1968 |
+
{
|
| 1969 |
+
"epoch": 0.7962185935999191,
|
| 1970 |
+
"grad_norm": 3.731058359146118,
|
| 1971 |
+
"learning_rate": 1.0207273531549181e-05,
|
| 1972 |
+
"loss": 4.2283,
|
| 1973 |
+
"step": 25200
|
| 1974 |
+
},
|
| 1975 |
+
{
|
| 1976 |
+
"epoch": 0.7993781911935696,
|
| 1977 |
+
"grad_norm": 4.568631172180176,
|
| 1978 |
+
"learning_rate": 1.0049290656892794e-05,
|
| 1979 |
+
"loss": 4.2678,
|
| 1980 |
+
"step": 25300
|
| 1981 |
+
},
|
| 1982 |
+
{
|
| 1983 |
+
"epoch": 0.80253778878722,
|
| 1984 |
+
"grad_norm": 4.449248313903809,
|
| 1985 |
+
"learning_rate": 9.891307782236406e-06,
|
| 1986 |
+
"loss": 4.2368,
|
| 1987 |
+
"step": 25400
|
| 1988 |
+
},
|
| 1989 |
+
{
|
| 1990 |
+
"epoch": 0.8056973863808705,
|
| 1991 |
+
"grad_norm": 4.026270389556885,
|
| 1992 |
+
"learning_rate": 9.733324907580019e-06,
|
| 1993 |
+
"loss": 4.2688,
|
| 1994 |
+
"step": 25500
|
| 1995 |
+
},
|
| 1996 |
+
{
|
| 1997 |
+
"epoch": 0.808856983974521,
|
| 1998 |
+
"grad_norm": 4.677060604095459,
|
| 1999 |
+
"learning_rate": 9.575342032923631e-06,
|
| 2000 |
+
"loss": 4.2312,
|
| 2001 |
+
"step": 25600
|
| 2002 |
+
},
|
| 2003 |
+
{
|
| 2004 |
+
"epoch": 0.8120165815681715,
|
| 2005 |
+
"grad_norm": 3.9102447032928467,
|
| 2006 |
+
"learning_rate": 9.417359158267244e-06,
|
| 2007 |
+
"loss": 4.2257,
|
| 2008 |
+
"step": 25700
|
| 2009 |
+
},
|
| 2010 |
+
{
|
| 2011 |
+
"epoch": 0.815176179161822,
|
| 2012 |
+
"grad_norm": 5.438532829284668,
|
| 2013 |
+
"learning_rate": 9.259376283610857e-06,
|
| 2014 |
+
"loss": 4.2285,
|
| 2015 |
+
"step": 25800
|
| 2016 |
+
},
|
| 2017 |
+
{
|
| 2018 |
+
"epoch": 0.8183357767554724,
|
| 2019 |
+
"grad_norm": 4.997466087341309,
|
| 2020 |
+
"learning_rate": 9.10139340895447e-06,
|
| 2021 |
+
"loss": 4.2642,
|
| 2022 |
+
"step": 25900
|
| 2023 |
+
},
|
| 2024 |
+
{
|
| 2025 |
+
"epoch": 0.8214953743491229,
|
| 2026 |
+
"grad_norm": 4.287134647369385,
|
| 2027 |
+
"learning_rate": 8.943410534298082e-06,
|
| 2028 |
+
"loss": 4.2341,
|
| 2029 |
+
"step": 26000
|
| 2030 |
+
},
|
| 2031 |
+
{
|
| 2032 |
+
"epoch": 0.8214953743491229,
|
| 2033 |
+
"eval_loss": 4.404691219329834,
|
| 2034 |
+
"eval_runtime": 706.9434,
|
| 2035 |
+
"eval_samples_per_second": 124.361,
|
| 2036 |
+
"eval_steps_per_second": 15.546,
|
| 2037 |
+
"step": 26000
|
| 2038 |
+
},
|
| 2039 |
+
{
|
| 2040 |
+
"epoch": 0.8246549719427734,
|
| 2041 |
+
"grad_norm": 4.450557708740234,
|
| 2042 |
+
"learning_rate": 8.785427659641696e-06,
|
| 2043 |
+
"loss": 4.289,
|
| 2044 |
+
"step": 26100
|
| 2045 |
+
},
|
| 2046 |
+
{
|
| 2047 |
+
"epoch": 0.8278145695364238,
|
| 2048 |
+
"grad_norm": 3.8313915729522705,
|
| 2049 |
+
"learning_rate": 8.627444784985309e-06,
|
| 2050 |
+
"loss": 4.2494,
|
| 2051 |
+
"step": 26200
|
| 2052 |
+
},
|
| 2053 |
+
{
|
| 2054 |
+
"epoch": 0.8309741671300743,
|
| 2055 |
+
"grad_norm": 4.760807514190674,
|
| 2056 |
+
"learning_rate": 8.469461910328922e-06,
|
| 2057 |
+
"loss": 4.2273,
|
| 2058 |
+
"step": 26300
|
| 2059 |
+
},
|
| 2060 |
+
{
|
| 2061 |
+
"epoch": 0.8341337647237248,
|
| 2062 |
+
"grad_norm": 4.606013774871826,
|
| 2063 |
+
"learning_rate": 8.311479035672534e-06,
|
| 2064 |
+
"loss": 4.2665,
|
| 2065 |
+
"step": 26400
|
| 2066 |
+
},
|
| 2067 |
+
{
|
| 2068 |
+
"epoch": 0.8372933623173753,
|
| 2069 |
+
"grad_norm": 3.884476661682129,
|
| 2070 |
+
"learning_rate": 8.153496161016145e-06,
|
| 2071 |
+
"loss": 4.2543,
|
| 2072 |
+
"step": 26500
|
| 2073 |
+
},
|
| 2074 |
+
{
|
| 2075 |
+
"epoch": 0.8404529599110258,
|
| 2076 |
+
"grad_norm": 4.1962127685546875,
|
| 2077 |
+
"learning_rate": 7.995513286359758e-06,
|
| 2078 |
+
"loss": 4.2865,
|
| 2079 |
+
"step": 26600
|
| 2080 |
+
},
|
| 2081 |
+
{
|
| 2082 |
+
"epoch": 0.8436125575046762,
|
| 2083 |
+
"grad_norm": 4.536484718322754,
|
| 2084 |
+
"learning_rate": 7.83753041170337e-06,
|
| 2085 |
+
"loss": 4.2786,
|
| 2086 |
+
"step": 26700
|
| 2087 |
+
},
|
| 2088 |
+
{
|
| 2089 |
+
"epoch": 0.8467721550983267,
|
| 2090 |
+
"grad_norm": 4.506146430969238,
|
| 2091 |
+
"learning_rate": 7.679547537046985e-06,
|
| 2092 |
+
"loss": 4.276,
|
| 2093 |
+
"step": 26800
|
| 2094 |
+
},
|
| 2095 |
+
{
|
| 2096 |
+
"epoch": 0.8499317526919772,
|
| 2097 |
+
"grad_norm": 5.557046413421631,
|
| 2098 |
+
"learning_rate": 7.521564662390597e-06,
|
| 2099 |
+
"loss": 4.2689,
|
| 2100 |
+
"step": 26900
|
| 2101 |
+
},
|
| 2102 |
+
{
|
| 2103 |
+
"epoch": 0.8530913502856277,
|
| 2104 |
+
"grad_norm": 4.260463237762451,
|
| 2105 |
+
"learning_rate": 7.365161616480774e-06,
|
| 2106 |
+
"loss": 4.2908,
|
| 2107 |
+
"step": 27000
|
| 2108 |
+
},
|
| 2109 |
+
{
|
| 2110 |
+
"epoch": 0.8530913502856277,
|
| 2111 |
+
"eval_loss": 4.402127265930176,
|
| 2112 |
+
"eval_runtime": 706.1851,
|
| 2113 |
+
"eval_samples_per_second": 124.494,
|
| 2114 |
+
"eval_steps_per_second": 15.562,
|
| 2115 |
+
"step": 27000
|
| 2116 |
+
},
|
| 2117 |
+
{
|
| 2118 |
+
"epoch": 0.8562509478792781,
|
| 2119 |
+
"grad_norm": 3.9850101470947266,
|
| 2120 |
+
"learning_rate": 7.207178741824388e-06,
|
| 2121 |
+
"loss": 4.2581,
|
| 2122 |
+
"step": 27100
|
| 2123 |
+
},
|
| 2124 |
+
{
|
| 2125 |
+
"epoch": 0.8594105454729286,
|
| 2126 |
+
"grad_norm": 5.228881359100342,
|
| 2127 |
+
"learning_rate": 7.049195867168e-06,
|
| 2128 |
+
"loss": 4.2347,
|
| 2129 |
+
"step": 27200
|
| 2130 |
+
},
|
| 2131 |
+
{
|
| 2132 |
+
"epoch": 0.8625701430665791,
|
| 2133 |
+
"grad_norm": 4.810471534729004,
|
| 2134 |
+
"learning_rate": 6.891212992511611e-06,
|
| 2135 |
+
"loss": 4.2821,
|
| 2136 |
+
"step": 27300
|
| 2137 |
+
},
|
| 2138 |
+
{
|
| 2139 |
+
"epoch": 0.8657297406602296,
|
| 2140 |
+
"grad_norm": 5.416518688201904,
|
| 2141 |
+
"learning_rate": 6.733230117855225e-06,
|
| 2142 |
+
"loss": 4.3175,
|
| 2143 |
+
"step": 27400
|
| 2144 |
+
},
|
| 2145 |
+
{
|
| 2146 |
+
"epoch": 0.86888933825388,
|
| 2147 |
+
"grad_norm": 4.266486167907715,
|
| 2148 |
+
"learning_rate": 6.575247243198837e-06,
|
| 2149 |
+
"loss": 4.219,
|
| 2150 |
+
"step": 27500
|
| 2151 |
+
},
|
| 2152 |
+
{
|
| 2153 |
+
"epoch": 0.8720489358475304,
|
| 2154 |
+
"grad_norm": 3.9196479320526123,
|
| 2155 |
+
"learning_rate": 6.41726436854245e-06,
|
| 2156 |
+
"loss": 4.2921,
|
| 2157 |
+
"step": 27600
|
| 2158 |
+
},
|
| 2159 |
+
{
|
| 2160 |
+
"epoch": 0.8752085334411809,
|
| 2161 |
+
"grad_norm": 3.8927292823791504,
|
| 2162 |
+
"learning_rate": 6.2592814938860626e-06,
|
| 2163 |
+
"loss": 4.224,
|
| 2164 |
+
"step": 27700
|
| 2165 |
+
},
|
| 2166 |
+
{
|
| 2167 |
+
"epoch": 0.8783681310348314,
|
| 2168 |
+
"grad_norm": 4.370504379272461,
|
| 2169 |
+
"learning_rate": 6.101298619229675e-06,
|
| 2170 |
+
"loss": 4.2426,
|
| 2171 |
+
"step": 27800
|
| 2172 |
+
},
|
| 2173 |
+
{
|
| 2174 |
+
"epoch": 0.8815277286284818,
|
| 2175 |
+
"grad_norm": 4.366906642913818,
|
| 2176 |
+
"learning_rate": 5.943315744573289e-06,
|
| 2177 |
+
"loss": 4.2668,
|
| 2178 |
+
"step": 27900
|
| 2179 |
+
},
|
| 2180 |
+
{
|
| 2181 |
+
"epoch": 0.8846873262221323,
|
| 2182 |
+
"grad_norm": 3.702138662338257,
|
| 2183 |
+
"learning_rate": 5.785332869916901e-06,
|
| 2184 |
+
"loss": 4.2331,
|
| 2185 |
+
"step": 28000
|
| 2186 |
+
},
|
| 2187 |
+
{
|
| 2188 |
+
"epoch": 0.8846873262221323,
|
| 2189 |
+
"eval_loss": 4.399970054626465,
|
| 2190 |
+
"eval_runtime": 705.6917,
|
| 2191 |
+
"eval_samples_per_second": 124.581,
|
| 2192 |
+
"eval_steps_per_second": 15.573,
|
| 2193 |
+
"step": 28000
|
| 2194 |
+
},
|
| 2195 |
+
{
|
| 2196 |
+
"epoch": 0.8878469238157828,
|
| 2197 |
+
"grad_norm": 3.696819543838501,
|
| 2198 |
+
"learning_rate": 5.627349995260514e-06,
|
| 2199 |
+
"loss": 4.2842,
|
| 2200 |
+
"step": 28100
|
| 2201 |
+
},
|
| 2202 |
+
{
|
| 2203 |
+
"epoch": 0.8910065214094333,
|
| 2204 |
+
"grad_norm": 3.837357759475708,
|
| 2205 |
+
"learning_rate": 5.469367120604127e-06,
|
| 2206 |
+
"loss": 4.2652,
|
| 2207 |
+
"step": 28200
|
| 2208 |
+
},
|
| 2209 |
+
{
|
| 2210 |
+
"epoch": 0.8941661190030837,
|
| 2211 |
+
"grad_norm": 4.607182502746582,
|
| 2212 |
+
"learning_rate": 5.31138424594774e-06,
|
| 2213 |
+
"loss": 4.2412,
|
| 2214 |
+
"step": 28300
|
| 2215 |
+
},
|
| 2216 |
+
{
|
| 2217 |
+
"epoch": 0.8973257165967342,
|
| 2218 |
+
"grad_norm": 3.803236484527588,
|
| 2219 |
+
"learning_rate": 5.154981200037916e-06,
|
| 2220 |
+
"loss": 4.2347,
|
| 2221 |
+
"step": 28400
|
| 2222 |
+
},
|
| 2223 |
+
{
|
| 2224 |
+
"epoch": 0.9004853141903847,
|
| 2225 |
+
"grad_norm": 4.182679653167725,
|
| 2226 |
+
"learning_rate": 4.9969983253815286e-06,
|
| 2227 |
+
"loss": 4.2815,
|
| 2228 |
+
"step": 28500
|
| 2229 |
+
},
|
| 2230 |
+
{
|
| 2231 |
+
"epoch": 0.9036449117840352,
|
| 2232 |
+
"grad_norm": 4.844381809234619,
|
| 2233 |
+
"learning_rate": 4.839015450725141e-06,
|
| 2234 |
+
"loss": 4.2565,
|
| 2235 |
+
"step": 28600
|
| 2236 |
+
},
|
| 2237 |
+
{
|
| 2238 |
+
"epoch": 0.9068045093776856,
|
| 2239 |
+
"grad_norm": 4.660447597503662,
|
| 2240 |
+
"learning_rate": 4.681032576068755e-06,
|
| 2241 |
+
"loss": 4.2364,
|
| 2242 |
+
"step": 28700
|
| 2243 |
+
},
|
| 2244 |
+
{
|
| 2245 |
+
"epoch": 0.9099641069713361,
|
| 2246 |
+
"grad_norm": 4.16555643081665,
|
| 2247 |
+
"learning_rate": 4.523049701412367e-06,
|
| 2248 |
+
"loss": 4.2775,
|
| 2249 |
+
"step": 28800
|
| 2250 |
+
},
|
| 2251 |
+
{
|
| 2252 |
+
"epoch": 0.9131237045649866,
|
| 2253 |
+
"grad_norm": 4.574762344360352,
|
| 2254 |
+
"learning_rate": 4.36506682675598e-06,
|
| 2255 |
+
"loss": 4.2255,
|
| 2256 |
+
"step": 28900
|
| 2257 |
+
},
|
| 2258 |
+
{
|
| 2259 |
+
"epoch": 0.9162833021586371,
|
| 2260 |
+
"grad_norm": 5.035285949707031,
|
| 2261 |
+
"learning_rate": 4.207083952099593e-06,
|
| 2262 |
+
"loss": 4.2787,
|
| 2263 |
+
"step": 29000
|
| 2264 |
+
},
|
| 2265 |
+
{
|
| 2266 |
+
"epoch": 0.9162833021586371,
|
| 2267 |
+
"eval_loss": 4.398350238800049,
|
| 2268 |
+
"eval_runtime": 702.9123,
|
| 2269 |
+
"eval_samples_per_second": 125.074,
|
| 2270 |
+
"eval_steps_per_second": 15.635,
|
| 2271 |
+
"step": 29000
|
| 2272 |
+
},
|
| 2273 |
+
{
|
| 2274 |
+
"epoch": 0.9194428997522875,
|
| 2275 |
+
"grad_norm": 4.223283290863037,
|
| 2276 |
+
"learning_rate": 4.049101077443205e-06,
|
| 2277 |
+
"loss": 4.2364,
|
| 2278 |
+
"step": 29100
|
| 2279 |
+
},
|
| 2280 |
+
{
|
| 2281 |
+
"epoch": 0.922602497345938,
|
| 2282 |
+
"grad_norm": 4.202021598815918,
|
| 2283 |
+
"learning_rate": 3.891118202786819e-06,
|
| 2284 |
+
"loss": 4.2258,
|
| 2285 |
+
"step": 29200
|
| 2286 |
+
},
|
| 2287 |
+
{
|
| 2288 |
+
"epoch": 0.9257620949395885,
|
| 2289 |
+
"grad_norm": 4.703751087188721,
|
| 2290 |
+
"learning_rate": 3.7331353281304305e-06,
|
| 2291 |
+
"loss": 4.2532,
|
| 2292 |
+
"step": 29300
|
| 2293 |
+
},
|
| 2294 |
+
{
|
| 2295 |
+
"epoch": 0.928921692533239,
|
| 2296 |
+
"grad_norm": 4.034895420074463,
|
| 2297 |
+
"learning_rate": 3.5751524534740436e-06,
|
| 2298 |
+
"loss": 4.2942,
|
| 2299 |
+
"step": 29400
|
| 2300 |
+
},
|
| 2301 |
+
{
|
| 2302 |
+
"epoch": 0.9320812901268895,
|
| 2303 |
+
"grad_norm": 5.5384521484375,
|
| 2304 |
+
"learning_rate": 3.417169578817656e-06,
|
| 2305 |
+
"loss": 4.1981,
|
| 2306 |
+
"step": 29500
|
| 2307 |
+
},
|
| 2308 |
+
{
|
| 2309 |
+
"epoch": 0.9352408877205399,
|
| 2310 |
+
"grad_norm": 4.258658409118652,
|
| 2311 |
+
"learning_rate": 3.2591867041612693e-06,
|
| 2312 |
+
"loss": 4.2552,
|
| 2313 |
+
"step": 29600
|
| 2314 |
+
},
|
| 2315 |
+
{
|
| 2316 |
+
"epoch": 0.9384004853141904,
|
| 2317 |
+
"grad_norm": 4.754565715789795,
|
| 2318 |
+
"learning_rate": 3.101203829504882e-06,
|
| 2319 |
+
"loss": 4.2769,
|
| 2320 |
+
"step": 29700
|
| 2321 |
+
},
|
| 2322 |
+
{
|
| 2323 |
+
"epoch": 0.9415600829078409,
|
| 2324 |
+
"grad_norm": 4.568323612213135,
|
| 2325 |
+
"learning_rate": 2.9432209548484945e-06,
|
| 2326 |
+
"loss": 4.2332,
|
| 2327 |
+
"step": 29800
|
| 2328 |
+
},
|
| 2329 |
+
{
|
| 2330 |
+
"epoch": 0.9447196805014914,
|
| 2331 |
+
"grad_norm": 4.62890625,
|
| 2332 |
+
"learning_rate": 2.785238080192107e-06,
|
| 2333 |
+
"loss": 4.2182,
|
| 2334 |
+
"step": 29900
|
| 2335 |
+
},
|
| 2336 |
+
{
|
| 2337 |
+
"epoch": 0.9478792780951418,
|
| 2338 |
+
"grad_norm": 4.388423919677734,
|
| 2339 |
+
"learning_rate": 2.6272552055357202e-06,
|
| 2340 |
+
"loss": 4.2405,
|
| 2341 |
+
"step": 30000
|
| 2342 |
+
},
|
| 2343 |
+
{
|
| 2344 |
+
"epoch": 0.9478792780951418,
|
| 2345 |
+
"eval_loss": 4.397016525268555,
|
| 2346 |
+
"eval_runtime": 702.9258,
|
| 2347 |
+
"eval_samples_per_second": 125.072,
|
| 2348 |
+
"eval_steps_per_second": 15.635,
|
| 2349 |
+
"step": 30000
|
| 2350 |
+
}
|
| 2351 |
+
],
|
| 2352 |
+
"logging_steps": 100,
|
| 2353 |
+
"max_steps": 31649,
|
| 2354 |
+
"num_input_tokens_seen": 0,
|
| 2355 |
+
"num_train_epochs": 1,
|
| 2356 |
+
"save_steps": 10000,
|
| 2357 |
+
"stateful_callbacks": {
|
| 2358 |
+
"TrainerControl": {
|
| 2359 |
+
"args": {
|
| 2360 |
+
"should_epoch_stop": false,
|
| 2361 |
+
"should_evaluate": false,
|
| 2362 |
+
"should_log": false,
|
| 2363 |
+
"should_save": true,
|
| 2364 |
+
"should_training_stop": false
|
| 2365 |
+
},
|
| 2366 |
+
"attributes": {}
|
| 2367 |
+
}
|
| 2368 |
+
},
|
| 2369 |
+
"total_flos": 1.37434595328e+17,
|
| 2370 |
+
"train_batch_size": 5,
|
| 2371 |
+
"trial_name": null,
|
| 2372 |
+
"trial_params": null
|
| 2373 |
+
}
|
fine_tuned_model_output_1/checkpoint-30000/training_args.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:95acebf5e801652183ac4ad483ff8bfe2ede739795ede7240bbee3b5a7707481
|
| 3 |
+
size 5368
|
fine_tuned_model_output_1/checkpoint-31649/config.json
ADDED
|
@@ -0,0 +1,30 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"_name_or_path": "./linked_model_with_custom_tokenizer",
|
| 3 |
+
"architectures": [
|
| 4 |
+
"Qwen2ForCausalLM"
|
| 5 |
+
],
|
| 6 |
+
"attention_dropout": 0.0,
|
| 7 |
+
"bos_token_id": 151643,
|
| 8 |
+
"eos_token_id": 151643,
|
| 9 |
+
"hidden_act": "silu",
|
| 10 |
+
"hidden_size": 896,
|
| 11 |
+
"initializer_range": 0.02,
|
| 12 |
+
"intermediate_size": 4864,
|
| 13 |
+
"max_position_embeddings": 32768,
|
| 14 |
+
"max_window_layers": 24,
|
| 15 |
+
"model_type": "qwen2",
|
| 16 |
+
"num_attention_heads": 14,
|
| 17 |
+
"num_hidden_layers": 4,
|
| 18 |
+
"num_key_value_heads": 2,
|
| 19 |
+
"rms_norm_eps": 1e-06,
|
| 20 |
+
"rope_scaling": null,
|
| 21 |
+
"rope_theta": 1000000.0,
|
| 22 |
+
"sliding_window": 32768,
|
| 23 |
+
"tie_word_embeddings": true,
|
| 24 |
+
"torch_dtype": "float32",
|
| 25 |
+
"transformers_version": "4.49.0",
|
| 26 |
+
"use_cache": true,
|
| 27 |
+
"use_mrope": false,
|
| 28 |
+
"use_sliding_window": false,
|
| 29 |
+
"vocab_size": 32002
|
| 30 |
+
}
|
fine_tuned_model_output_1/checkpoint-31649/generation_config.json
ADDED
|
@@ -0,0 +1,6 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"_from_model_config": true,
|
| 3 |
+
"bos_token_id": 151643,
|
| 4 |
+
"eos_token_id": 151643,
|
| 5 |
+
"transformers_version": "4.49.0"
|
| 6 |
+
}
|
fine_tuned_model_output_1/checkpoint-31649/model.safetensors
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:afcce6cf3c1847de57b3a7d784f76f5a4cd17e92a12f7e0d9c0b1e09a0c11289
|
| 3 |
+
size 353302384
|
fine_tuned_model_output_1/checkpoint-31649/optimizer.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:645c57c019bc35a30fa6e2cd34a6d48463c54b5a4c1b4a8ddec4358ffcbd270b
|
| 3 |
+
size 706636090
|
fine_tuned_model_output_1/checkpoint-31649/rng_state.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:7c73101b0d1b062977c6bc2c28e1818cec6565f6e118fbf6b45147e7aa88e76a
|
| 3 |
+
size 14244
|
fine_tuned_model_output_1/checkpoint-31649/scaler.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:ddafcbc8d0cb976e9945bb240001e5e941fa59171574e3bc96dad9fe2692a662
|
| 3 |
+
size 988
|
fine_tuned_model_output_1/checkpoint-31649/scheduler.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:cd52aa5084e929007ec458d89ce1911daabdd9c03e18e8c15fe1989c104af1e8
|
| 3 |
+
size 1064
|
fine_tuned_model_output_1/checkpoint-31649/special_tokens_map.json
ADDED
|
@@ -0,0 +1,30 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"bos_token": {
|
| 3 |
+
"content": "<s>",
|
| 4 |
+
"lstrip": false,
|
| 5 |
+
"normalized": false,
|
| 6 |
+
"rstrip": false,
|
| 7 |
+
"single_word": false
|
| 8 |
+
},
|
| 9 |
+
"eos_token": {
|
| 10 |
+
"content": "</s>",
|
| 11 |
+
"lstrip": false,
|
| 12 |
+
"normalized": false,
|
| 13 |
+
"rstrip": false,
|
| 14 |
+
"single_word": false
|
| 15 |
+
},
|
| 16 |
+
"pad_token": {
|
| 17 |
+
"content": "</s>",
|
| 18 |
+
"lstrip": false,
|
| 19 |
+
"normalized": false,
|
| 20 |
+
"rstrip": false,
|
| 21 |
+
"single_word": false
|
| 22 |
+
},
|
| 23 |
+
"unk_token": {
|
| 24 |
+
"content": "<unk>",
|
| 25 |
+
"lstrip": false,
|
| 26 |
+
"normalized": false,
|
| 27 |
+
"rstrip": false,
|
| 28 |
+
"single_word": false
|
| 29 |
+
}
|
| 30 |
+
}
|
fine_tuned_model_output_1/checkpoint-31649/tokenizer.json
ADDED
|
The diff for this file is too large to render.
See raw diff
|
|
|
fine_tuned_model_output_1/checkpoint-31649/tokenizer_config.json
ADDED
|
@@ -0,0 +1,67 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"added_tokens_decoder": {
|
| 3 |
+
"0": {
|
| 4 |
+
"content": "<unk>",
|
| 5 |
+
"lstrip": false,
|
| 6 |
+
"normalized": false,
|
| 7 |
+
"rstrip": false,
|
| 8 |
+
"single_word": false,
|
| 9 |
+
"special": true
|
| 10 |
+
},
|
| 11 |
+
"1": {
|
| 12 |
+
"content": "<s>",
|
| 13 |
+
"lstrip": false,
|
| 14 |
+
"normalized": false,
|
| 15 |
+
"rstrip": false,
|
| 16 |
+
"single_word": false,
|
| 17 |
+
"special": true
|
| 18 |
+
},
|
| 19 |
+
"2": {
|
| 20 |
+
"content": "</s>",
|
| 21 |
+
"lstrip": false,
|
| 22 |
+
"normalized": false,
|
| 23 |
+
"rstrip": false,
|
| 24 |
+
"single_word": false,
|
| 25 |
+
"special": true
|
| 26 |
+
},
|
| 27 |
+
"3": {
|
| 28 |
+
"content": "<pad>",
|
| 29 |
+
"lstrip": false,
|
| 30 |
+
"normalized": false,
|
| 31 |
+
"rstrip": false,
|
| 32 |
+
"single_word": false,
|
| 33 |
+
"special": true
|
| 34 |
+
},
|
| 35 |
+
"32000": {
|
| 36 |
+
"content": "Sanskrit:",
|
| 37 |
+
"lstrip": false,
|
| 38 |
+
"normalized": true,
|
| 39 |
+
"rstrip": false,
|
| 40 |
+
"single_word": false,
|
| 41 |
+
"special": false
|
| 42 |
+
},
|
| 43 |
+
"32001": {
|
| 44 |
+
"content": "English:",
|
| 45 |
+
"lstrip": false,
|
| 46 |
+
"normalized": true,
|
| 47 |
+
"rstrip": false,
|
| 48 |
+
"single_word": false,
|
| 49 |
+
"special": false
|
| 50 |
+
}
|
| 51 |
+
},
|
| 52 |
+
"bos_token": "<s>",
|
| 53 |
+
"clean_up_tokenization_spaces": false,
|
| 54 |
+
"eos_token": "</s>",
|
| 55 |
+
"extra_special_tokens": {},
|
| 56 |
+
"max_length": 512,
|
| 57 |
+
"model_max_length": 1000000000000000019884624838656,
|
| 58 |
+
"pad_to_multiple_of": null,
|
| 59 |
+
"pad_token": "</s>",
|
| 60 |
+
"pad_token_type_id": 0,
|
| 61 |
+
"padding_side": "right",
|
| 62 |
+
"stride": 0,
|
| 63 |
+
"tokenizer_class": "PreTrainedTokenizer",
|
| 64 |
+
"truncation_side": "right",
|
| 65 |
+
"truncation_strategy": "longest_first",
|
| 66 |
+
"unk_token": "<unk>"
|
| 67 |
+
}
|
fine_tuned_model_output_1/checkpoint-31649/trainer_state.json
ADDED
|
@@ -0,0 +1,2493 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": 4.3962178230285645,
|
| 3 |
+
"best_model_checkpoint": "./fine_tuned_model_output_1\\checkpoint-31000",
|
| 4 |
+
"epoch": 0.9999810424144381,
|
| 5 |
+
"eval_steps": 1000,
|
| 6 |
+
"global_step": 31649,
|
| 7 |
+
"is_hyper_param_search": false,
|
| 8 |
+
"is_local_process_zero": true,
|
| 9 |
+
"is_world_process_zero": true,
|
| 10 |
+
"log_history": [
|
| 11 |
+
{
|
| 12 |
+
"epoch": 0.0031595975936504726,
|
| 13 |
+
"grad_norm": 5.00691556930542,
|
| 14 |
+
"learning_rate": 4.984517678283674e-05,
|
| 15 |
+
"loss": 5.1637,
|
| 16 |
+
"step": 100
|
| 17 |
+
},
|
| 18 |
+
{
|
| 19 |
+
"epoch": 0.006319195187300945,
|
| 20 |
+
"grad_norm": 4.790140628814697,
|
| 21 |
+
"learning_rate": 4.968719390818036e-05,
|
| 22 |
+
"loss": 4.9551,
|
| 23 |
+
"step": 200
|
| 24 |
+
},
|
| 25 |
+
{
|
| 26 |
+
"epoch": 0.009478792780951418,
|
| 27 |
+
"grad_norm": 4.75383186340332,
|
| 28 |
+
"learning_rate": 4.952921103352397e-05,
|
| 29 |
+
"loss": 4.8786,
|
| 30 |
+
"step": 300
|
| 31 |
+
},
|
| 32 |
+
{
|
| 33 |
+
"epoch": 0.01263839037460189,
|
| 34 |
+
"grad_norm": 5.218325138092041,
|
| 35 |
+
"learning_rate": 4.9371228158867584e-05,
|
| 36 |
+
"loss": 4.8117,
|
| 37 |
+
"step": 400
|
| 38 |
+
},
|
| 39 |
+
{
|
| 40 |
+
"epoch": 0.015797987968252363,
|
| 41 |
+
"grad_norm": 5.101843357086182,
|
| 42 |
+
"learning_rate": 4.9213245284211194e-05,
|
| 43 |
+
"loss": 4.8153,
|
| 44 |
+
"step": 500
|
| 45 |
+
},
|
| 46 |
+
{
|
| 47 |
+
"epoch": 0.018957585561902836,
|
| 48 |
+
"grad_norm": 4.165743350982666,
|
| 49 |
+
"learning_rate": 4.905526240955481e-05,
|
| 50 |
+
"loss": 4.7331,
|
| 51 |
+
"step": 600
|
| 52 |
+
},
|
| 53 |
+
{
|
| 54 |
+
"epoch": 0.02211718315555331,
|
| 55 |
+
"grad_norm": 4.386114597320557,
|
| 56 |
+
"learning_rate": 4.889727953489842e-05,
|
| 57 |
+
"loss": 4.7982,
|
| 58 |
+
"step": 700
|
| 59 |
+
},
|
| 60 |
+
{
|
| 61 |
+
"epoch": 0.02527678074920378,
|
| 62 |
+
"grad_norm": 4.677374362945557,
|
| 63 |
+
"learning_rate": 4.8739296660242035e-05,
|
| 64 |
+
"loss": 4.751,
|
| 65 |
+
"step": 800
|
| 66 |
+
},
|
| 67 |
+
{
|
| 68 |
+
"epoch": 0.028436378342854254,
|
| 69 |
+
"grad_norm": 4.1365814208984375,
|
| 70 |
+
"learning_rate": 4.8581313785585644e-05,
|
| 71 |
+
"loss": 4.7631,
|
| 72 |
+
"step": 900
|
| 73 |
+
},
|
| 74 |
+
{
|
| 75 |
+
"epoch": 0.031595975936504726,
|
| 76 |
+
"grad_norm": 4.651081562042236,
|
| 77 |
+
"learning_rate": 4.842333091092926e-05,
|
| 78 |
+
"loss": 4.7504,
|
| 79 |
+
"step": 1000
|
| 80 |
+
},
|
| 81 |
+
{
|
| 82 |
+
"epoch": 0.031595975936504726,
|
| 83 |
+
"eval_loss": 4.873119354248047,
|
| 84 |
+
"eval_runtime": 665.3679,
|
| 85 |
+
"eval_samples_per_second": 132.131,
|
| 86 |
+
"eval_steps_per_second": 16.517,
|
| 87 |
+
"step": 1000
|
| 88 |
+
},
|
| 89 |
+
{
|
| 90 |
+
"epoch": 0.0347555735301552,
|
| 91 |
+
"grad_norm": 5.823278903961182,
|
| 92 |
+
"learning_rate": 4.826534803627287e-05,
|
| 93 |
+
"loss": 4.6999,
|
| 94 |
+
"step": 1100
|
| 95 |
+
},
|
| 96 |
+
{
|
| 97 |
+
"epoch": 0.03791517112380567,
|
| 98 |
+
"grad_norm": 4.6261773109436035,
|
| 99 |
+
"learning_rate": 4.8107365161616486e-05,
|
| 100 |
+
"loss": 4.6704,
|
| 101 |
+
"step": 1200
|
| 102 |
+
},
|
| 103 |
+
{
|
| 104 |
+
"epoch": 0.04107476871745615,
|
| 105 |
+
"grad_norm": 4.6412200927734375,
|
| 106 |
+
"learning_rate": 4.79493822869601e-05,
|
| 107 |
+
"loss": 4.691,
|
| 108 |
+
"step": 1300
|
| 109 |
+
},
|
| 110 |
+
{
|
| 111 |
+
"epoch": 0.04423436631110662,
|
| 112 |
+
"grad_norm": 4.368412494659424,
|
| 113 |
+
"learning_rate": 4.779139941230371e-05,
|
| 114 |
+
"loss": 4.7048,
|
| 115 |
+
"step": 1400
|
| 116 |
+
},
|
| 117 |
+
{
|
| 118 |
+
"epoch": 0.04739396390475709,
|
| 119 |
+
"grad_norm": 3.9410157203674316,
|
| 120 |
+
"learning_rate": 4.763499636639389e-05,
|
| 121 |
+
"loss": 4.6213,
|
| 122 |
+
"step": 1500
|
| 123 |
+
},
|
| 124 |
+
{
|
| 125 |
+
"epoch": 0.05055356149840756,
|
| 126 |
+
"grad_norm": 5.697903633117676,
|
| 127 |
+
"learning_rate": 4.74770134917375e-05,
|
| 128 |
+
"loss": 4.6791,
|
| 129 |
+
"step": 1600
|
| 130 |
+
},
|
| 131 |
+
{
|
| 132 |
+
"epoch": 0.05371315909205804,
|
| 133 |
+
"grad_norm": 4.182404041290283,
|
| 134 |
+
"learning_rate": 4.7319030617081114e-05,
|
| 135 |
+
"loss": 4.6493,
|
| 136 |
+
"step": 1700
|
| 137 |
+
},
|
| 138 |
+
{
|
| 139 |
+
"epoch": 0.05687275668570851,
|
| 140 |
+
"grad_norm": 3.8856332302093506,
|
| 141 |
+
"learning_rate": 4.7161047742424724e-05,
|
| 142 |
+
"loss": 4.6081,
|
| 143 |
+
"step": 1800
|
| 144 |
+
},
|
| 145 |
+
{
|
| 146 |
+
"epoch": 0.06003235427935898,
|
| 147 |
+
"grad_norm": 3.778503656387329,
|
| 148 |
+
"learning_rate": 4.700306486776834e-05,
|
| 149 |
+
"loss": 4.6437,
|
| 150 |
+
"step": 1900
|
| 151 |
+
},
|
| 152 |
+
{
|
| 153 |
+
"epoch": 0.06319195187300945,
|
| 154 |
+
"grad_norm": 5.072572231292725,
|
| 155 |
+
"learning_rate": 4.684508199311195e-05,
|
| 156 |
+
"loss": 4.5587,
|
| 157 |
+
"step": 2000
|
| 158 |
+
},
|
| 159 |
+
{
|
| 160 |
+
"epoch": 0.06319195187300945,
|
| 161 |
+
"eval_loss": 4.763667106628418,
|
| 162 |
+
"eval_runtime": 665.5888,
|
| 163 |
+
"eval_samples_per_second": 132.088,
|
| 164 |
+
"eval_steps_per_second": 16.512,
|
| 165 |
+
"step": 2000
|
| 166 |
+
},
|
| 167 |
+
{
|
| 168 |
+
"epoch": 0.06635154946665993,
|
| 169 |
+
"grad_norm": 5.395784854888916,
|
| 170 |
+
"learning_rate": 4.6687099118455565e-05,
|
| 171 |
+
"loss": 4.5621,
|
| 172 |
+
"step": 2100
|
| 173 |
+
},
|
| 174 |
+
{
|
| 175 |
+
"epoch": 0.0695111470603104,
|
| 176 |
+
"grad_norm": 4.856666088104248,
|
| 177 |
+
"learning_rate": 4.6529116243799174e-05,
|
| 178 |
+
"loss": 4.5721,
|
| 179 |
+
"step": 2200
|
| 180 |
+
},
|
| 181 |
+
{
|
| 182 |
+
"epoch": 0.07267074465396087,
|
| 183 |
+
"grad_norm": 4.935929298400879,
|
| 184 |
+
"learning_rate": 4.637113336914279e-05,
|
| 185 |
+
"loss": 4.6004,
|
| 186 |
+
"step": 2300
|
| 187 |
+
},
|
| 188 |
+
{
|
| 189 |
+
"epoch": 0.07583034224761134,
|
| 190 |
+
"grad_norm": 5.257358551025391,
|
| 191 |
+
"learning_rate": 4.62131504944864e-05,
|
| 192 |
+
"loss": 4.6032,
|
| 193 |
+
"step": 2400
|
| 194 |
+
},
|
| 195 |
+
{
|
| 196 |
+
"epoch": 0.07898993984126182,
|
| 197 |
+
"grad_norm": 4.67532205581665,
|
| 198 |
+
"learning_rate": 4.6055167619830016e-05,
|
| 199 |
+
"loss": 4.5734,
|
| 200 |
+
"step": 2500
|
| 201 |
+
},
|
| 202 |
+
{
|
| 203 |
+
"epoch": 0.0821495374349123,
|
| 204 |
+
"grad_norm": 4.023921966552734,
|
| 205 |
+
"learning_rate": 4.5897184745173625e-05,
|
| 206 |
+
"loss": 4.5888,
|
| 207 |
+
"step": 2600
|
| 208 |
+
},
|
| 209 |
+
{
|
| 210 |
+
"epoch": 0.08530913502856276,
|
| 211 |
+
"grad_norm": 3.9619414806365967,
|
| 212 |
+
"learning_rate": 4.573920187051724e-05,
|
| 213 |
+
"loss": 4.5703,
|
| 214 |
+
"step": 2700
|
| 215 |
+
},
|
| 216 |
+
{
|
| 217 |
+
"epoch": 0.08846873262221323,
|
| 218 |
+
"grad_norm": 3.7340047359466553,
|
| 219 |
+
"learning_rate": 4.558121899586085e-05,
|
| 220 |
+
"loss": 4.5437,
|
| 221 |
+
"step": 2800
|
| 222 |
+
},
|
| 223 |
+
{
|
| 224 |
+
"epoch": 0.09162833021586371,
|
| 225 |
+
"grad_norm": 3.5358519554138184,
|
| 226 |
+
"learning_rate": 4.5423236121204466e-05,
|
| 227 |
+
"loss": 4.5433,
|
| 228 |
+
"step": 2900
|
| 229 |
+
},
|
| 230 |
+
{
|
| 231 |
+
"epoch": 0.09478792780951419,
|
| 232 |
+
"grad_norm": 5.275386810302734,
|
| 233 |
+
"learning_rate": 4.5265253246548075e-05,
|
| 234 |
+
"loss": 4.5315,
|
| 235 |
+
"step": 3000
|
| 236 |
+
},
|
| 237 |
+
{
|
| 238 |
+
"epoch": 0.09478792780951419,
|
| 239 |
+
"eval_loss": 4.6958231925964355,
|
| 240 |
+
"eval_runtime": 660.1374,
|
| 241 |
+
"eval_samples_per_second": 133.178,
|
| 242 |
+
"eval_steps_per_second": 16.648,
|
| 243 |
+
"step": 3000
|
| 244 |
+
},
|
| 245 |
+
{
|
| 246 |
+
"epoch": 0.09794752540316465,
|
| 247 |
+
"grad_norm": 4.195945739746094,
|
| 248 |
+
"learning_rate": 4.510727037189169e-05,
|
| 249 |
+
"loss": 4.5292,
|
| 250 |
+
"step": 3100
|
| 251 |
+
},
|
| 252 |
+
{
|
| 253 |
+
"epoch": 0.10110712299681512,
|
| 254 |
+
"grad_norm": 4.976147651672363,
|
| 255 |
+
"learning_rate": 4.49492874972353e-05,
|
| 256 |
+
"loss": 4.5718,
|
| 257 |
+
"step": 3200
|
| 258 |
+
},
|
| 259 |
+
{
|
| 260 |
+
"epoch": 0.1042667205904656,
|
| 261 |
+
"grad_norm": 3.999481678009033,
|
| 262 |
+
"learning_rate": 4.479130462257892e-05,
|
| 263 |
+
"loss": 4.5177,
|
| 264 |
+
"step": 3300
|
| 265 |
+
},
|
| 266 |
+
{
|
| 267 |
+
"epoch": 0.10742631818411608,
|
| 268 |
+
"grad_norm": 5.00895357131958,
|
| 269 |
+
"learning_rate": 4.463332174792253e-05,
|
| 270 |
+
"loss": 4.5418,
|
| 271 |
+
"step": 3400
|
| 272 |
+
},
|
| 273 |
+
{
|
| 274 |
+
"epoch": 0.11058591577776654,
|
| 275 |
+
"grad_norm": 4.499265193939209,
|
| 276 |
+
"learning_rate": 4.447533887326614e-05,
|
| 277 |
+
"loss": 4.5262,
|
| 278 |
+
"step": 3500
|
| 279 |
+
},
|
| 280 |
+
{
|
| 281 |
+
"epoch": 0.11374551337141701,
|
| 282 |
+
"grad_norm": 4.686629772186279,
|
| 283 |
+
"learning_rate": 4.431735599860976e-05,
|
| 284 |
+
"loss": 4.5276,
|
| 285 |
+
"step": 3600
|
| 286 |
+
},
|
| 287 |
+
{
|
| 288 |
+
"epoch": 0.11690511096506749,
|
| 289 |
+
"grad_norm": 5.327017784118652,
|
| 290 |
+
"learning_rate": 4.415937312395337e-05,
|
| 291 |
+
"loss": 4.4939,
|
| 292 |
+
"step": 3700
|
| 293 |
+
},
|
| 294 |
+
{
|
| 295 |
+
"epoch": 0.12006470855871797,
|
| 296 |
+
"grad_norm": 3.8358302116394043,
|
| 297 |
+
"learning_rate": 4.400139024929698e-05,
|
| 298 |
+
"loss": 4.5028,
|
| 299 |
+
"step": 3800
|
| 300 |
+
},
|
| 301 |
+
{
|
| 302 |
+
"epoch": 0.12322430615236843,
|
| 303 |
+
"grad_norm": 4.16530704498291,
|
| 304 |
+
"learning_rate": 4.384340737464059e-05,
|
| 305 |
+
"loss": 4.4964,
|
| 306 |
+
"step": 3900
|
| 307 |
+
},
|
| 308 |
+
{
|
| 309 |
+
"epoch": 0.1263839037460189,
|
| 310 |
+
"grad_norm": 4.142209053039551,
|
| 311 |
+
"learning_rate": 4.36854244999842e-05,
|
| 312 |
+
"loss": 4.4899,
|
| 313 |
+
"step": 4000
|
| 314 |
+
},
|
| 315 |
+
{
|
| 316 |
+
"epoch": 0.1263839037460189,
|
| 317 |
+
"eval_loss": 4.649657249450684,
|
| 318 |
+
"eval_runtime": 689.9636,
|
| 319 |
+
"eval_samples_per_second": 127.421,
|
| 320 |
+
"eval_steps_per_second": 15.928,
|
| 321 |
+
"step": 4000
|
| 322 |
+
},
|
| 323 |
+
{
|
| 324 |
+
"epoch": 0.12954350133966938,
|
| 325 |
+
"grad_norm": 3.615328311920166,
|
| 326 |
+
"learning_rate": 4.352744162532781e-05,
|
| 327 |
+
"loss": 4.4671,
|
| 328 |
+
"step": 4100
|
| 329 |
+
},
|
| 330 |
+
{
|
| 331 |
+
"epoch": 0.13270309893331986,
|
| 332 |
+
"grad_norm": 5.139493465423584,
|
| 333 |
+
"learning_rate": 4.336945875067143e-05,
|
| 334 |
+
"loss": 4.4954,
|
| 335 |
+
"step": 4200
|
| 336 |
+
},
|
| 337 |
+
{
|
| 338 |
+
"epoch": 0.13586269652697033,
|
| 339 |
+
"grad_norm": 4.381932258605957,
|
| 340 |
+
"learning_rate": 4.3211475876015036e-05,
|
| 341 |
+
"loss": 4.5103,
|
| 342 |
+
"step": 4300
|
| 343 |
+
},
|
| 344 |
+
{
|
| 345 |
+
"epoch": 0.1390222941206208,
|
| 346 |
+
"grad_norm": 4.257575988769531,
|
| 347 |
+
"learning_rate": 4.305349300135865e-05,
|
| 348 |
+
"loss": 4.4677,
|
| 349 |
+
"step": 4400
|
| 350 |
+
},
|
| 351 |
+
{
|
| 352 |
+
"epoch": 0.14218189171427126,
|
| 353 |
+
"grad_norm": 3.5805318355560303,
|
| 354 |
+
"learning_rate": 4.289551012670227e-05,
|
| 355 |
+
"loss": 4.4547,
|
| 356 |
+
"step": 4500
|
| 357 |
+
},
|
| 358 |
+
{
|
| 359 |
+
"epoch": 0.14534148930792173,
|
| 360 |
+
"grad_norm": 4.602927207946777,
|
| 361 |
+
"learning_rate": 4.273752725204588e-05,
|
| 362 |
+
"loss": 4.4885,
|
| 363 |
+
"step": 4600
|
| 364 |
+
},
|
| 365 |
+
{
|
| 366 |
+
"epoch": 0.1485010869015722,
|
| 367 |
+
"grad_norm": 4.054294586181641,
|
| 368 |
+
"learning_rate": 4.2579544377389494e-05,
|
| 369 |
+
"loss": 4.4791,
|
| 370 |
+
"step": 4700
|
| 371 |
+
},
|
| 372 |
+
{
|
| 373 |
+
"epoch": 0.1516606844952227,
|
| 374 |
+
"grad_norm": 3.857389450073242,
|
| 375 |
+
"learning_rate": 4.24215615027331e-05,
|
| 376 |
+
"loss": 4.4316,
|
| 377 |
+
"step": 4800
|
| 378 |
+
},
|
| 379 |
+
{
|
| 380 |
+
"epoch": 0.15482028208887316,
|
| 381 |
+
"grad_norm": 3.913726806640625,
|
| 382 |
+
"learning_rate": 4.226357862807672e-05,
|
| 383 |
+
"loss": 4.4786,
|
| 384 |
+
"step": 4900
|
| 385 |
+
},
|
| 386 |
+
{
|
| 387 |
+
"epoch": 0.15797987968252364,
|
| 388 |
+
"grad_norm": 4.159733295440674,
|
| 389 |
+
"learning_rate": 4.210559575342033e-05,
|
| 390 |
+
"loss": 4.422,
|
| 391 |
+
"step": 5000
|
| 392 |
+
},
|
| 393 |
+
{
|
| 394 |
+
"epoch": 0.15797987968252364,
|
| 395 |
+
"eval_loss": 4.6132493019104,
|
| 396 |
+
"eval_runtime": 721.543,
|
| 397 |
+
"eval_samples_per_second": 121.844,
|
| 398 |
+
"eval_steps_per_second": 15.231,
|
| 399 |
+
"step": 5000
|
| 400 |
+
},
|
| 401 |
+
{
|
| 402 |
+
"epoch": 0.16113947727617411,
|
| 403 |
+
"grad_norm": 5.403946876525879,
|
| 404 |
+
"learning_rate": 4.1947612878763944e-05,
|
| 405 |
+
"loss": 4.4857,
|
| 406 |
+
"step": 5100
|
| 407 |
+
},
|
| 408 |
+
{
|
| 409 |
+
"epoch": 0.1642990748698246,
|
| 410 |
+
"grad_norm": 11.68526554107666,
|
| 411 |
+
"learning_rate": 4.178963000410755e-05,
|
| 412 |
+
"loss": 4.4777,
|
| 413 |
+
"step": 5200
|
| 414 |
+
},
|
| 415 |
+
{
|
| 416 |
+
"epoch": 0.16745867246347504,
|
| 417 |
+
"grad_norm": 4.937869548797607,
|
| 418 |
+
"learning_rate": 4.163164712945117e-05,
|
| 419 |
+
"loss": 4.4452,
|
| 420 |
+
"step": 5300
|
| 421 |
+
},
|
| 422 |
+
{
|
| 423 |
+
"epoch": 0.17061827005712552,
|
| 424 |
+
"grad_norm": 3.7432169914245605,
|
| 425 |
+
"learning_rate": 4.147366425479478e-05,
|
| 426 |
+
"loss": 4.4717,
|
| 427 |
+
"step": 5400
|
| 428 |
+
},
|
| 429 |
+
{
|
| 430 |
+
"epoch": 0.173777867650776,
|
| 431 |
+
"grad_norm": 3.801788806915283,
|
| 432 |
+
"learning_rate": 4.1317261208884964e-05,
|
| 433 |
+
"loss": 4.4424,
|
| 434 |
+
"step": 5500
|
| 435 |
+
},
|
| 436 |
+
{
|
| 437 |
+
"epoch": 0.17693746524442647,
|
| 438 |
+
"grad_norm": 4.984846115112305,
|
| 439 |
+
"learning_rate": 4.115927833422857e-05,
|
| 440 |
+
"loss": 4.4671,
|
| 441 |
+
"step": 5600
|
| 442 |
+
},
|
| 443 |
+
{
|
| 444 |
+
"epoch": 0.18009706283807694,
|
| 445 |
+
"grad_norm": 3.8125479221343994,
|
| 446 |
+
"learning_rate": 4.100287528831875e-05,
|
| 447 |
+
"loss": 4.4079,
|
| 448 |
+
"step": 5700
|
| 449 |
+
},
|
| 450 |
+
{
|
| 451 |
+
"epoch": 0.18325666043172742,
|
| 452 |
+
"grad_norm": 5.528467655181885,
|
| 453 |
+
"learning_rate": 4.084489241366236e-05,
|
| 454 |
+
"loss": 4.4194,
|
| 455 |
+
"step": 5800
|
| 456 |
+
},
|
| 457 |
+
{
|
| 458 |
+
"epoch": 0.1864162580253779,
|
| 459 |
+
"grad_norm": 4.0301899909973145,
|
| 460 |
+
"learning_rate": 4.068690953900598e-05,
|
| 461 |
+
"loss": 4.4299,
|
| 462 |
+
"step": 5900
|
| 463 |
+
},
|
| 464 |
+
{
|
| 465 |
+
"epoch": 0.18957585561902837,
|
| 466 |
+
"grad_norm": 3.4992246627807617,
|
| 467 |
+
"learning_rate": 4.0528926664349586e-05,
|
| 468 |
+
"loss": 4.448,
|
| 469 |
+
"step": 6000
|
| 470 |
+
},
|
| 471 |
+
{
|
| 472 |
+
"epoch": 0.18957585561902837,
|
| 473 |
+
"eval_loss": 4.5857086181640625,
|
| 474 |
+
"eval_runtime": 702.1155,
|
| 475 |
+
"eval_samples_per_second": 125.216,
|
| 476 |
+
"eval_steps_per_second": 15.653,
|
| 477 |
+
"step": 6000
|
| 478 |
+
},
|
| 479 |
+
{
|
| 480 |
+
"epoch": 0.19273545321267882,
|
| 481 |
+
"grad_norm": 4.121975421905518,
|
| 482 |
+
"learning_rate": 4.03709437896932e-05,
|
| 483 |
+
"loss": 4.4076,
|
| 484 |
+
"step": 6100
|
| 485 |
+
},
|
| 486 |
+
{
|
| 487 |
+
"epoch": 0.1958950508063293,
|
| 488 |
+
"grad_norm": 3.662929058074951,
|
| 489 |
+
"learning_rate": 4.021296091503681e-05,
|
| 490 |
+
"loss": 4.3953,
|
| 491 |
+
"step": 6200
|
| 492 |
+
},
|
| 493 |
+
{
|
| 494 |
+
"epoch": 0.19905464839997977,
|
| 495 |
+
"grad_norm": 4.008944988250732,
|
| 496 |
+
"learning_rate": 4.005497804038043e-05,
|
| 497 |
+
"loss": 4.4581,
|
| 498 |
+
"step": 6300
|
| 499 |
+
},
|
| 500 |
+
{
|
| 501 |
+
"epoch": 0.20221424599363025,
|
| 502 |
+
"grad_norm": 3.7961764335632324,
|
| 503 |
+
"learning_rate": 3.9896995165724036e-05,
|
| 504 |
+
"loss": 4.397,
|
| 505 |
+
"step": 6400
|
| 506 |
+
},
|
| 507 |
+
{
|
| 508 |
+
"epoch": 0.20537384358728072,
|
| 509 |
+
"grad_norm": 4.710026741027832,
|
| 510 |
+
"learning_rate": 3.973901229106765e-05,
|
| 511 |
+
"loss": 4.4515,
|
| 512 |
+
"step": 6500
|
| 513 |
+
},
|
| 514 |
+
{
|
| 515 |
+
"epoch": 0.2085334411809312,
|
| 516 |
+
"grad_norm": 4.220251083374023,
|
| 517 |
+
"learning_rate": 3.958102941641126e-05,
|
| 518 |
+
"loss": 4.4199,
|
| 519 |
+
"step": 6600
|
| 520 |
+
},
|
| 521 |
+
{
|
| 522 |
+
"epoch": 0.21169303877458168,
|
| 523 |
+
"grad_norm": 3.7243406772613525,
|
| 524 |
+
"learning_rate": 3.942304654175488e-05,
|
| 525 |
+
"loss": 4.4097,
|
| 526 |
+
"step": 6700
|
| 527 |
+
},
|
| 528 |
+
{
|
| 529 |
+
"epoch": 0.21485263636823215,
|
| 530 |
+
"grad_norm": 4.340923309326172,
|
| 531 |
+
"learning_rate": 3.926664349584505e-05,
|
| 532 |
+
"loss": 4.4105,
|
| 533 |
+
"step": 6800
|
| 534 |
+
},
|
| 535 |
+
{
|
| 536 |
+
"epoch": 0.2180122339618826,
|
| 537 |
+
"grad_norm": 4.304684638977051,
|
| 538 |
+
"learning_rate": 3.9108660621188665e-05,
|
| 539 |
+
"loss": 4.4033,
|
| 540 |
+
"step": 6900
|
| 541 |
+
},
|
| 542 |
+
{
|
| 543 |
+
"epoch": 0.22117183155553308,
|
| 544 |
+
"grad_norm": 4.587896347045898,
|
| 545 |
+
"learning_rate": 3.8950677746532275e-05,
|
| 546 |
+
"loss": 4.4451,
|
| 547 |
+
"step": 7000
|
| 548 |
+
},
|
| 549 |
+
{
|
| 550 |
+
"epoch": 0.22117183155553308,
|
| 551 |
+
"eval_loss": 4.5615153312683105,
|
| 552 |
+
"eval_runtime": 695.974,
|
| 553 |
+
"eval_samples_per_second": 126.321,
|
| 554 |
+
"eval_steps_per_second": 15.791,
|
| 555 |
+
"step": 7000
|
| 556 |
+
},
|
| 557 |
+
{
|
| 558 |
+
"epoch": 0.22433142914918355,
|
| 559 |
+
"grad_norm": 4.658904075622559,
|
| 560 |
+
"learning_rate": 3.879269487187589e-05,
|
| 561 |
+
"loss": 4.407,
|
| 562 |
+
"step": 7100
|
| 563 |
+
},
|
| 564 |
+
{
|
| 565 |
+
"epoch": 0.22749102674283403,
|
| 566 |
+
"grad_norm": 4.226349830627441,
|
| 567 |
+
"learning_rate": 3.86347119972195e-05,
|
| 568 |
+
"loss": 4.4055,
|
| 569 |
+
"step": 7200
|
| 570 |
+
},
|
| 571 |
+
{
|
| 572 |
+
"epoch": 0.2306506243364845,
|
| 573 |
+
"grad_norm": 5.705229759216309,
|
| 574 |
+
"learning_rate": 3.8476729122563116e-05,
|
| 575 |
+
"loss": 4.3862,
|
| 576 |
+
"step": 7300
|
| 577 |
+
},
|
| 578 |
+
{
|
| 579 |
+
"epoch": 0.23381022193013498,
|
| 580 |
+
"grad_norm": 3.9644031524658203,
|
| 581 |
+
"learning_rate": 3.8318746247906725e-05,
|
| 582 |
+
"loss": 4.3647,
|
| 583 |
+
"step": 7400
|
| 584 |
+
},
|
| 585 |
+
{
|
| 586 |
+
"epoch": 0.23696981952378546,
|
| 587 |
+
"grad_norm": 4.256899356842041,
|
| 588 |
+
"learning_rate": 3.816076337325034e-05,
|
| 589 |
+
"loss": 4.4091,
|
| 590 |
+
"step": 7500
|
| 591 |
+
},
|
| 592 |
+
{
|
| 593 |
+
"epoch": 0.24012941711743593,
|
| 594 |
+
"grad_norm": 3.951305627822876,
|
| 595 |
+
"learning_rate": 3.800278049859396e-05,
|
| 596 |
+
"loss": 4.3994,
|
| 597 |
+
"step": 7600
|
| 598 |
+
},
|
| 599 |
+
{
|
| 600 |
+
"epoch": 0.2432890147110864,
|
| 601 |
+
"grad_norm": 3.6965837478637695,
|
| 602 |
+
"learning_rate": 3.7844797623937566e-05,
|
| 603 |
+
"loss": 4.4043,
|
| 604 |
+
"step": 7700
|
| 605 |
+
},
|
| 606 |
+
{
|
| 607 |
+
"epoch": 0.24644861230473686,
|
| 608 |
+
"grad_norm": 4.412446022033691,
|
| 609 |
+
"learning_rate": 3.768681474928118e-05,
|
| 610 |
+
"loss": 4.4237,
|
| 611 |
+
"step": 7800
|
| 612 |
+
},
|
| 613 |
+
{
|
| 614 |
+
"epoch": 0.24960820989838733,
|
| 615 |
+
"grad_norm": 4.190601348876953,
|
| 616 |
+
"learning_rate": 3.752883187462479e-05,
|
| 617 |
+
"loss": 4.365,
|
| 618 |
+
"step": 7900
|
| 619 |
+
},
|
| 620 |
+
{
|
| 621 |
+
"epoch": 0.2527678074920378,
|
| 622 |
+
"grad_norm": 3.6600217819213867,
|
| 623 |
+
"learning_rate": 3.737084899996841e-05,
|
| 624 |
+
"loss": 4.393,
|
| 625 |
+
"step": 8000
|
| 626 |
+
},
|
| 627 |
+
{
|
| 628 |
+
"epoch": 0.2527678074920378,
|
| 629 |
+
"eval_loss": 4.542718887329102,
|
| 630 |
+
"eval_runtime": 692.645,
|
| 631 |
+
"eval_samples_per_second": 126.928,
|
| 632 |
+
"eval_steps_per_second": 15.867,
|
| 633 |
+
"step": 8000
|
| 634 |
+
},
|
| 635 |
+
{
|
| 636 |
+
"epoch": 0.2559274050856883,
|
| 637 |
+
"grad_norm": 6.242320537567139,
|
| 638 |
+
"learning_rate": 3.721286612531202e-05,
|
| 639 |
+
"loss": 4.4185,
|
| 640 |
+
"step": 8100
|
| 641 |
+
},
|
| 642 |
+
{
|
| 643 |
+
"epoch": 0.25908700267933876,
|
| 644 |
+
"grad_norm": 4.762045383453369,
|
| 645 |
+
"learning_rate": 3.705488325065563e-05,
|
| 646 |
+
"loss": 4.3656,
|
| 647 |
+
"step": 8200
|
| 648 |
+
},
|
| 649 |
+
{
|
| 650 |
+
"epoch": 0.26224660027298924,
|
| 651 |
+
"grad_norm": 5.066886901855469,
|
| 652 |
+
"learning_rate": 3.689690037599924e-05,
|
| 653 |
+
"loss": 4.394,
|
| 654 |
+
"step": 8300
|
| 655 |
+
},
|
| 656 |
+
{
|
| 657 |
+
"epoch": 0.2654061978666397,
|
| 658 |
+
"grad_norm": 4.975124835968018,
|
| 659 |
+
"learning_rate": 3.673891750134286e-05,
|
| 660 |
+
"loss": 4.3346,
|
| 661 |
+
"step": 8400
|
| 662 |
+
},
|
| 663 |
+
{
|
| 664 |
+
"epoch": 0.2685657954602902,
|
| 665 |
+
"grad_norm": 4.436885356903076,
|
| 666 |
+
"learning_rate": 3.658093462668647e-05,
|
| 667 |
+
"loss": 4.3958,
|
| 668 |
+
"step": 8500
|
| 669 |
+
},
|
| 670 |
+
{
|
| 671 |
+
"epoch": 0.27172539305394067,
|
| 672 |
+
"grad_norm": 4.618080139160156,
|
| 673 |
+
"learning_rate": 3.6422951752030084e-05,
|
| 674 |
+
"loss": 4.3791,
|
| 675 |
+
"step": 8600
|
| 676 |
+
},
|
| 677 |
+
{
|
| 678 |
+
"epoch": 0.27488499064759114,
|
| 679 |
+
"grad_norm": 4.878961086273193,
|
| 680 |
+
"learning_rate": 3.626496887737369e-05,
|
| 681 |
+
"loss": 4.3738,
|
| 682 |
+
"step": 8700
|
| 683 |
+
},
|
| 684 |
+
{
|
| 685 |
+
"epoch": 0.2780445882412416,
|
| 686 |
+
"grad_norm": 3.7712392807006836,
|
| 687 |
+
"learning_rate": 3.610698600271731e-05,
|
| 688 |
+
"loss": 4.362,
|
| 689 |
+
"step": 8800
|
| 690 |
+
},
|
| 691 |
+
{
|
| 692 |
+
"epoch": 0.2812041858348921,
|
| 693 |
+
"grad_norm": 4.318520545959473,
|
| 694 |
+
"learning_rate": 3.594900312806092e-05,
|
| 695 |
+
"loss": 4.367,
|
| 696 |
+
"step": 8900
|
| 697 |
+
},
|
| 698 |
+
{
|
| 699 |
+
"epoch": 0.2843637834285425,
|
| 700 |
+
"grad_norm": 3.3742518424987793,
|
| 701 |
+
"learning_rate": 3.5791020253404534e-05,
|
| 702 |
+
"loss": 4.3763,
|
| 703 |
+
"step": 9000
|
| 704 |
+
},
|
| 705 |
+
{
|
| 706 |
+
"epoch": 0.2843637834285425,
|
| 707 |
+
"eval_loss": 4.525923728942871,
|
| 708 |
+
"eval_runtime": 692.5212,
|
| 709 |
+
"eval_samples_per_second": 126.951,
|
| 710 |
+
"eval_steps_per_second": 15.87,
|
| 711 |
+
"step": 9000
|
| 712 |
+
},
|
| 713 |
+
{
|
| 714 |
+
"epoch": 0.287523381022193,
|
| 715 |
+
"grad_norm": 4.027231693267822,
|
| 716 |
+
"learning_rate": 3.5633037378748143e-05,
|
| 717 |
+
"loss": 4.4086,
|
| 718 |
+
"step": 9100
|
| 719 |
+
},
|
| 720 |
+
{
|
| 721 |
+
"epoch": 0.29068297861584347,
|
| 722 |
+
"grad_norm": 4.556132793426514,
|
| 723 |
+
"learning_rate": 3.547505450409176e-05,
|
| 724 |
+
"loss": 4.3686,
|
| 725 |
+
"step": 9200
|
| 726 |
+
},
|
| 727 |
+
{
|
| 728 |
+
"epoch": 0.29384257620949394,
|
| 729 |
+
"grad_norm": 3.9203240871429443,
|
| 730 |
+
"learning_rate": 3.531707162943537e-05,
|
| 731 |
+
"loss": 4.365,
|
| 732 |
+
"step": 9300
|
| 733 |
+
},
|
| 734 |
+
{
|
| 735 |
+
"epoch": 0.2970021738031444,
|
| 736 |
+
"grad_norm": 4.094367027282715,
|
| 737 |
+
"learning_rate": 3.5159088754778985e-05,
|
| 738 |
+
"loss": 4.3734,
|
| 739 |
+
"step": 9400
|
| 740 |
+
},
|
| 741 |
+
{
|
| 742 |
+
"epoch": 0.3001617713967949,
|
| 743 |
+
"grad_norm": 5.018619060516357,
|
| 744 |
+
"learning_rate": 3.50011058801226e-05,
|
| 745 |
+
"loss": 4.4014,
|
| 746 |
+
"step": 9500
|
| 747 |
+
},
|
| 748 |
+
{
|
| 749 |
+
"epoch": 0.3033213689904454,
|
| 750 |
+
"grad_norm": 3.928473711013794,
|
| 751 |
+
"learning_rate": 3.484312300546621e-05,
|
| 752 |
+
"loss": 4.3508,
|
| 753 |
+
"step": 9600
|
| 754 |
+
},
|
| 755 |
+
{
|
| 756 |
+
"epoch": 0.30648096658409585,
|
| 757 |
+
"grad_norm": 3.889299154281616,
|
| 758 |
+
"learning_rate": 3.4685140130809826e-05,
|
| 759 |
+
"loss": 4.365,
|
| 760 |
+
"step": 9700
|
| 761 |
+
},
|
| 762 |
+
{
|
| 763 |
+
"epoch": 0.3096405641777463,
|
| 764 |
+
"grad_norm": 4.932437896728516,
|
| 765 |
+
"learning_rate": 3.4527157256153435e-05,
|
| 766 |
+
"loss": 4.3166,
|
| 767 |
+
"step": 9800
|
| 768 |
+
},
|
| 769 |
+
{
|
| 770 |
+
"epoch": 0.3128001617713968,
|
| 771 |
+
"grad_norm": 4.175967216491699,
|
| 772 |
+
"learning_rate": 3.436917438149705e-05,
|
| 773 |
+
"loss": 4.3463,
|
| 774 |
+
"step": 9900
|
| 775 |
+
},
|
| 776 |
+
{
|
| 777 |
+
"epoch": 0.3159597593650473,
|
| 778 |
+
"grad_norm": 5.54374361038208,
|
| 779 |
+
"learning_rate": 3.421119150684066e-05,
|
| 780 |
+
"loss": 4.3506,
|
| 781 |
+
"step": 10000
|
| 782 |
+
},
|
| 783 |
+
{
|
| 784 |
+
"epoch": 0.3159597593650473,
|
| 785 |
+
"eval_loss": 4.511664390563965,
|
| 786 |
+
"eval_runtime": 691.9773,
|
| 787 |
+
"eval_samples_per_second": 127.05,
|
| 788 |
+
"eval_steps_per_second": 15.882,
|
| 789 |
+
"step": 10000
|
| 790 |
+
},
|
| 791 |
+
{
|
| 792 |
+
"epoch": 0.31911935695869775,
|
| 793 |
+
"grad_norm": 4.466663360595703,
|
| 794 |
+
"learning_rate": 3.4053208632184277e-05,
|
| 795 |
+
"loss": 4.3351,
|
| 796 |
+
"step": 10100
|
| 797 |
+
},
|
| 798 |
+
{
|
| 799 |
+
"epoch": 0.32227895455234823,
|
| 800 |
+
"grad_norm": 4.300159454345703,
|
| 801 |
+
"learning_rate": 3.3895225757527886e-05,
|
| 802 |
+
"loss": 4.3605,
|
| 803 |
+
"step": 10200
|
| 804 |
+
},
|
| 805 |
+
{
|
| 806 |
+
"epoch": 0.3254385521459987,
|
| 807 |
+
"grad_norm": 4.595406532287598,
|
| 808 |
+
"learning_rate": 3.37372428828715e-05,
|
| 809 |
+
"loss": 4.3432,
|
| 810 |
+
"step": 10300
|
| 811 |
+
},
|
| 812 |
+
{
|
| 813 |
+
"epoch": 0.3285981497396492,
|
| 814 |
+
"grad_norm": 4.077473163604736,
|
| 815 |
+
"learning_rate": 3.357926000821511e-05,
|
| 816 |
+
"loss": 4.3421,
|
| 817 |
+
"step": 10400
|
| 818 |
+
},
|
| 819 |
+
{
|
| 820 |
+
"epoch": 0.33175774733329966,
|
| 821 |
+
"grad_norm": 4.250525951385498,
|
| 822 |
+
"learning_rate": 3.342127713355873e-05,
|
| 823 |
+
"loss": 4.3795,
|
| 824 |
+
"step": 10500
|
| 825 |
+
},
|
| 826 |
+
{
|
| 827 |
+
"epoch": 0.3349173449269501,
|
| 828 |
+
"grad_norm": 3.9417407512664795,
|
| 829 |
+
"learning_rate": 3.3263294258902336e-05,
|
| 830 |
+
"loss": 4.3394,
|
| 831 |
+
"step": 10600
|
| 832 |
+
},
|
| 833 |
+
{
|
| 834 |
+
"epoch": 0.33807694252060055,
|
| 835 |
+
"grad_norm": 3.782409906387329,
|
| 836 |
+
"learning_rate": 3.310531138424595e-05,
|
| 837 |
+
"loss": 4.3061,
|
| 838 |
+
"step": 10700
|
| 839 |
+
},
|
| 840 |
+
{
|
| 841 |
+
"epoch": 0.34123654011425103,
|
| 842 |
+
"grad_norm": 4.497501850128174,
|
| 843 |
+
"learning_rate": 3.294732850958956e-05,
|
| 844 |
+
"loss": 4.3558,
|
| 845 |
+
"step": 10800
|
| 846 |
+
},
|
| 847 |
+
{
|
| 848 |
+
"epoch": 0.3443961377079015,
|
| 849 |
+
"grad_norm": 4.938557147979736,
|
| 850 |
+
"learning_rate": 3.278934563493318e-05,
|
| 851 |
+
"loss": 4.33,
|
| 852 |
+
"step": 10900
|
| 853 |
+
},
|
| 854 |
+
{
|
| 855 |
+
"epoch": 0.347555735301552,
|
| 856 |
+
"grad_norm": 4.101445198059082,
|
| 857 |
+
"learning_rate": 3.263136276027679e-05,
|
| 858 |
+
"loss": 4.3658,
|
| 859 |
+
"step": 11000
|
| 860 |
+
},
|
| 861 |
+
{
|
| 862 |
+
"epoch": 0.347555735301552,
|
| 863 |
+
"eval_loss": 4.499057769775391,
|
| 864 |
+
"eval_runtime": 696.9678,
|
| 865 |
+
"eval_samples_per_second": 126.141,
|
| 866 |
+
"eval_steps_per_second": 15.768,
|
| 867 |
+
"step": 11000
|
| 868 |
+
},
|
| 869 |
+
{
|
| 870 |
+
"epoch": 0.35071533289520246,
|
| 871 |
+
"grad_norm": 4.350194454193115,
|
| 872 |
+
"learning_rate": 3.24733798856204e-05,
|
| 873 |
+
"loss": 4.3129,
|
| 874 |
+
"step": 11100
|
| 875 |
+
},
|
| 876 |
+
{
|
| 877 |
+
"epoch": 0.35387493048885293,
|
| 878 |
+
"grad_norm": 4.244603157043457,
|
| 879 |
+
"learning_rate": 3.231539701096401e-05,
|
| 880 |
+
"loss": 4.3566,
|
| 881 |
+
"step": 11200
|
| 882 |
+
},
|
| 883 |
+
{
|
| 884 |
+
"epoch": 0.3570345280825034,
|
| 885 |
+
"grad_norm": 3.947141647338867,
|
| 886 |
+
"learning_rate": 3.215741413630762e-05,
|
| 887 |
+
"loss": 4.3572,
|
| 888 |
+
"step": 11300
|
| 889 |
+
},
|
| 890 |
+
{
|
| 891 |
+
"epoch": 0.3601941256761539,
|
| 892 |
+
"grad_norm": 4.241100311279297,
|
| 893 |
+
"learning_rate": 3.199943126165124e-05,
|
| 894 |
+
"loss": 4.3341,
|
| 895 |
+
"step": 11400
|
| 896 |
+
},
|
| 897 |
+
{
|
| 898 |
+
"epoch": 0.36335372326980436,
|
| 899 |
+
"grad_norm": 4.596353530883789,
|
| 900 |
+
"learning_rate": 3.184144838699485e-05,
|
| 901 |
+
"loss": 4.3702,
|
| 902 |
+
"step": 11500
|
| 903 |
+
},
|
| 904 |
+
{
|
| 905 |
+
"epoch": 0.36651332086345484,
|
| 906 |
+
"grad_norm": 4.667087554931641,
|
| 907 |
+
"learning_rate": 3.168346551233846e-05,
|
| 908 |
+
"loss": 4.3919,
|
| 909 |
+
"step": 11600
|
| 910 |
+
},
|
| 911 |
+
{
|
| 912 |
+
"epoch": 0.3696729184571053,
|
| 913 |
+
"grad_norm": 4.700224876403809,
|
| 914 |
+
"learning_rate": 3.152548263768207e-05,
|
| 915 |
+
"loss": 4.3622,
|
| 916 |
+
"step": 11700
|
| 917 |
+
},
|
| 918 |
+
{
|
| 919 |
+
"epoch": 0.3728325160507558,
|
| 920 |
+
"grad_norm": 4.907840728759766,
|
| 921 |
+
"learning_rate": 3.136749976302569e-05,
|
| 922 |
+
"loss": 4.3517,
|
| 923 |
+
"step": 11800
|
| 924 |
+
},
|
| 925 |
+
{
|
| 926 |
+
"epoch": 0.37599211364440627,
|
| 927 |
+
"grad_norm": 4.013961315155029,
|
| 928 |
+
"learning_rate": 3.12095168883693e-05,
|
| 929 |
+
"loss": 4.3636,
|
| 930 |
+
"step": 11900
|
| 931 |
+
},
|
| 932 |
+
{
|
| 933 |
+
"epoch": 0.37915171123805674,
|
| 934 |
+
"grad_norm": 3.6914377212524414,
|
| 935 |
+
"learning_rate": 3.105153401371291e-05,
|
| 936 |
+
"loss": 4.3087,
|
| 937 |
+
"step": 12000
|
| 938 |
+
},
|
| 939 |
+
{
|
| 940 |
+
"epoch": 0.37915171123805674,
|
| 941 |
+
"eval_loss": 4.48746395111084,
|
| 942 |
+
"eval_runtime": 715.6219,
|
| 943 |
+
"eval_samples_per_second": 122.853,
|
| 944 |
+
"eval_steps_per_second": 15.357,
|
| 945 |
+
"step": 12000
|
| 946 |
+
},
|
| 947 |
+
{
|
| 948 |
+
"epoch": 0.3823113088317072,
|
| 949 |
+
"grad_norm": 3.8867859840393066,
|
| 950 |
+
"learning_rate": 3.089355113905652e-05,
|
| 951 |
+
"loss": 4.337,
|
| 952 |
+
"step": 12100
|
| 953 |
+
},
|
| 954 |
+
{
|
| 955 |
+
"epoch": 0.38547090642535764,
|
| 956 |
+
"grad_norm": 4.648858070373535,
|
| 957 |
+
"learning_rate": 3.073556826440014e-05,
|
| 958 |
+
"loss": 4.3516,
|
| 959 |
+
"step": 12200
|
| 960 |
+
},
|
| 961 |
+
{
|
| 962 |
+
"epoch": 0.3886305040190081,
|
| 963 |
+
"grad_norm": 4.28641939163208,
|
| 964 |
+
"learning_rate": 3.0577585389743755e-05,
|
| 965 |
+
"loss": 4.3053,
|
| 966 |
+
"step": 12300
|
| 967 |
+
},
|
| 968 |
+
{
|
| 969 |
+
"epoch": 0.3917901016126586,
|
| 970 |
+
"grad_norm": 3.784064292907715,
|
| 971 |
+
"learning_rate": 3.0419602515087364e-05,
|
| 972 |
+
"loss": 4.3373,
|
| 973 |
+
"step": 12400
|
| 974 |
+
},
|
| 975 |
+
{
|
| 976 |
+
"epoch": 0.39494969920630907,
|
| 977 |
+
"grad_norm": 3.9344184398651123,
|
| 978 |
+
"learning_rate": 3.0261619640430976e-05,
|
| 979 |
+
"loss": 4.2861,
|
| 980 |
+
"step": 12500
|
| 981 |
+
},
|
| 982 |
+
{
|
| 983 |
+
"epoch": 0.39810929679995954,
|
| 984 |
+
"grad_norm": 4.234601974487305,
|
| 985 |
+
"learning_rate": 3.010363676577459e-05,
|
| 986 |
+
"loss": 4.3603,
|
| 987 |
+
"step": 12600
|
| 988 |
+
},
|
| 989 |
+
{
|
| 990 |
+
"epoch": 0.40126889439361,
|
| 991 |
+
"grad_norm": 4.8835835456848145,
|
| 992 |
+
"learning_rate": 2.9945653891118202e-05,
|
| 993 |
+
"loss": 4.3435,
|
| 994 |
+
"step": 12700
|
| 995 |
+
},
|
| 996 |
+
{
|
| 997 |
+
"epoch": 0.4044284919872605,
|
| 998 |
+
"grad_norm": 4.3561882972717285,
|
| 999 |
+
"learning_rate": 2.9787671016461814e-05,
|
| 1000 |
+
"loss": 4.3272,
|
| 1001 |
+
"step": 12800
|
| 1002 |
+
},
|
| 1003 |
+
{
|
| 1004 |
+
"epoch": 0.407588089580911,
|
| 1005 |
+
"grad_norm": 4.034718036651611,
|
| 1006 |
+
"learning_rate": 2.9629688141805427e-05,
|
| 1007 |
+
"loss": 4.2928,
|
| 1008 |
+
"step": 12900
|
| 1009 |
+
},
|
| 1010 |
+
{
|
| 1011 |
+
"epoch": 0.41074768717456145,
|
| 1012 |
+
"grad_norm": 4.587557315826416,
|
| 1013 |
+
"learning_rate": 2.947170526714904e-05,
|
| 1014 |
+
"loss": 4.3284,
|
| 1015 |
+
"step": 13000
|
| 1016 |
+
},
|
| 1017 |
+
{
|
| 1018 |
+
"epoch": 0.41074768717456145,
|
| 1019 |
+
"eval_loss": 4.477302074432373,
|
| 1020 |
+
"eval_runtime": 710.4818,
|
| 1021 |
+
"eval_samples_per_second": 123.741,
|
| 1022 |
+
"eval_steps_per_second": 15.468,
|
| 1023 |
+
"step": 13000
|
| 1024 |
+
},
|
| 1025 |
+
{
|
| 1026 |
+
"epoch": 0.4139072847682119,
|
| 1027 |
+
"grad_norm": 3.6614646911621094,
|
| 1028 |
+
"learning_rate": 2.9313722392492656e-05,
|
| 1029 |
+
"loss": 4.3301,
|
| 1030 |
+
"step": 13100
|
| 1031 |
+
},
|
| 1032 |
+
{
|
| 1033 |
+
"epoch": 0.4170668823618624,
|
| 1034 |
+
"grad_norm": 4.2207112312316895,
|
| 1035 |
+
"learning_rate": 2.915573951783627e-05,
|
| 1036 |
+
"loss": 4.3184,
|
| 1037 |
+
"step": 13200
|
| 1038 |
+
},
|
| 1039 |
+
{
|
| 1040 |
+
"epoch": 0.4202264799555129,
|
| 1041 |
+
"grad_norm": 4.571977138519287,
|
| 1042 |
+
"learning_rate": 2.8999336471926447e-05,
|
| 1043 |
+
"loss": 4.2802,
|
| 1044 |
+
"step": 13300
|
| 1045 |
+
},
|
| 1046 |
+
{
|
| 1047 |
+
"epoch": 0.42338607754916335,
|
| 1048 |
+
"grad_norm": 4.995270252227783,
|
| 1049 |
+
"learning_rate": 2.884135359727006e-05,
|
| 1050 |
+
"loss": 4.3286,
|
| 1051 |
+
"step": 13400
|
| 1052 |
+
},
|
| 1053 |
+
{
|
| 1054 |
+
"epoch": 0.42654567514281383,
|
| 1055 |
+
"grad_norm": 3.6203622817993164,
|
| 1056 |
+
"learning_rate": 2.8683370722613672e-05,
|
| 1057 |
+
"loss": 4.3566,
|
| 1058 |
+
"step": 13500
|
| 1059 |
+
},
|
| 1060 |
+
{
|
| 1061 |
+
"epoch": 0.4297052727364643,
|
| 1062 |
+
"grad_norm": 3.7246546745300293,
|
| 1063 |
+
"learning_rate": 2.852538784795728e-05,
|
| 1064 |
+
"loss": 4.3191,
|
| 1065 |
+
"step": 13600
|
| 1066 |
+
},
|
| 1067 |
+
{
|
| 1068 |
+
"epoch": 0.4328648703301148,
|
| 1069 |
+
"grad_norm": 4.118934631347656,
|
| 1070 |
+
"learning_rate": 2.8367404973300894e-05,
|
| 1071 |
+
"loss": 4.3338,
|
| 1072 |
+
"step": 13700
|
| 1073 |
+
},
|
| 1074 |
+
{
|
| 1075 |
+
"epoch": 0.4360244679237652,
|
| 1076 |
+
"grad_norm": 4.569579601287842,
|
| 1077 |
+
"learning_rate": 2.8209422098644506e-05,
|
| 1078 |
+
"loss": 4.3436,
|
| 1079 |
+
"step": 13800
|
| 1080 |
+
},
|
| 1081 |
+
{
|
| 1082 |
+
"epoch": 0.4391840655174157,
|
| 1083 |
+
"grad_norm": 4.280219554901123,
|
| 1084 |
+
"learning_rate": 2.805143922398812e-05,
|
| 1085 |
+
"loss": 4.2796,
|
| 1086 |
+
"step": 13900
|
| 1087 |
+
},
|
| 1088 |
+
{
|
| 1089 |
+
"epoch": 0.44234366311106615,
|
| 1090 |
+
"grad_norm": 4.866397857666016,
|
| 1091 |
+
"learning_rate": 2.7893456349331732e-05,
|
| 1092 |
+
"loss": 4.314,
|
| 1093 |
+
"step": 14000
|
| 1094 |
+
},
|
| 1095 |
+
{
|
| 1096 |
+
"epoch": 0.44234366311106615,
|
| 1097 |
+
"eval_loss": 4.468830585479736,
|
| 1098 |
+
"eval_runtime": 707.6161,
|
| 1099 |
+
"eval_samples_per_second": 124.243,
|
| 1100 |
+
"eval_steps_per_second": 15.531,
|
| 1101 |
+
"step": 14000
|
| 1102 |
+
},
|
| 1103 |
+
{
|
| 1104 |
+
"epoch": 0.44550326070471663,
|
| 1105 |
+
"grad_norm": 4.688275337219238,
|
| 1106 |
+
"learning_rate": 2.7735473474675344e-05,
|
| 1107 |
+
"loss": 4.2977,
|
| 1108 |
+
"step": 14100
|
| 1109 |
+
},
|
| 1110 |
+
{
|
| 1111 |
+
"epoch": 0.4486628582983671,
|
| 1112 |
+
"grad_norm": 4.715358734130859,
|
| 1113 |
+
"learning_rate": 2.7577490600018957e-05,
|
| 1114 |
+
"loss": 4.357,
|
| 1115 |
+
"step": 14200
|
| 1116 |
+
},
|
| 1117 |
+
{
|
| 1118 |
+
"epoch": 0.4518224558920176,
|
| 1119 |
+
"grad_norm": 4.193299293518066,
|
| 1120 |
+
"learning_rate": 2.741950772536257e-05,
|
| 1121 |
+
"loss": 4.3002,
|
| 1122 |
+
"step": 14300
|
| 1123 |
+
},
|
| 1124 |
+
{
|
| 1125 |
+
"epoch": 0.45498205348566806,
|
| 1126 |
+
"grad_norm": 5.159686088562012,
|
| 1127 |
+
"learning_rate": 2.7261524850706182e-05,
|
| 1128 |
+
"loss": 4.2935,
|
| 1129 |
+
"step": 14400
|
| 1130 |
+
},
|
| 1131 |
+
{
|
| 1132 |
+
"epoch": 0.45814165107931853,
|
| 1133 |
+
"grad_norm": 4.735175609588623,
|
| 1134 |
+
"learning_rate": 2.7103541976049795e-05,
|
| 1135 |
+
"loss": 4.3297,
|
| 1136 |
+
"step": 14500
|
| 1137 |
+
},
|
| 1138 |
+
{
|
| 1139 |
+
"epoch": 0.461301248672969,
|
| 1140 |
+
"grad_norm": 4.604960918426514,
|
| 1141 |
+
"learning_rate": 2.6945559101393408e-05,
|
| 1142 |
+
"loss": 4.308,
|
| 1143 |
+
"step": 14600
|
| 1144 |
+
},
|
| 1145 |
+
{
|
| 1146 |
+
"epoch": 0.4644608462666195,
|
| 1147 |
+
"grad_norm": 4.662102222442627,
|
| 1148 |
+
"learning_rate": 2.678757622673702e-05,
|
| 1149 |
+
"loss": 4.2911,
|
| 1150 |
+
"step": 14700
|
| 1151 |
+
},
|
| 1152 |
+
{
|
| 1153 |
+
"epoch": 0.46762044386026996,
|
| 1154 |
+
"grad_norm": 5.6145477294921875,
|
| 1155 |
+
"learning_rate": 2.6629593352080633e-05,
|
| 1156 |
+
"loss": 4.283,
|
| 1157 |
+
"step": 14800
|
| 1158 |
+
},
|
| 1159 |
+
{
|
| 1160 |
+
"epoch": 0.47078004145392044,
|
| 1161 |
+
"grad_norm": 5.1714606285095215,
|
| 1162 |
+
"learning_rate": 2.6471610477424246e-05,
|
| 1163 |
+
"loss": 4.3117,
|
| 1164 |
+
"step": 14900
|
| 1165 |
+
},
|
| 1166 |
+
{
|
| 1167 |
+
"epoch": 0.4739396390475709,
|
| 1168 |
+
"grad_norm": 4.684764385223389,
|
| 1169 |
+
"learning_rate": 2.6313627602767858e-05,
|
| 1170 |
+
"loss": 4.3159,
|
| 1171 |
+
"step": 15000
|
| 1172 |
+
},
|
| 1173 |
+
{
|
| 1174 |
+
"epoch": 0.4739396390475709,
|
| 1175 |
+
"eval_loss": 4.460094451904297,
|
| 1176 |
+
"eval_runtime": 708.6547,
|
| 1177 |
+
"eval_samples_per_second": 124.06,
|
| 1178 |
+
"eval_steps_per_second": 15.508,
|
| 1179 |
+
"step": 15000
|
| 1180 |
+
},
|
| 1181 |
+
{
|
| 1182 |
+
"epoch": 0.4770992366412214,
|
| 1183 |
+
"grad_norm": 4.812125205993652,
|
| 1184 |
+
"learning_rate": 2.6155644728111474e-05,
|
| 1185 |
+
"loss": 4.297,
|
| 1186 |
+
"step": 15100
|
| 1187 |
+
},
|
| 1188 |
+
{
|
| 1189 |
+
"epoch": 0.48025883423487187,
|
| 1190 |
+
"grad_norm": 4.402428150177002,
|
| 1191 |
+
"learning_rate": 2.5997661853455087e-05,
|
| 1192 |
+
"loss": 4.2707,
|
| 1193 |
+
"step": 15200
|
| 1194 |
+
},
|
| 1195 |
+
{
|
| 1196 |
+
"epoch": 0.48341843182852234,
|
| 1197 |
+
"grad_norm": 3.6931350231170654,
|
| 1198 |
+
"learning_rate": 2.5841258807545265e-05,
|
| 1199 |
+
"loss": 4.2843,
|
| 1200 |
+
"step": 15300
|
| 1201 |
+
},
|
| 1202 |
+
{
|
| 1203 |
+
"epoch": 0.4865780294221728,
|
| 1204 |
+
"grad_norm": 4.092090129852295,
|
| 1205 |
+
"learning_rate": 2.5683275932888878e-05,
|
| 1206 |
+
"loss": 4.2672,
|
| 1207 |
+
"step": 15400
|
| 1208 |
+
},
|
| 1209 |
+
{
|
| 1210 |
+
"epoch": 0.48973762701582324,
|
| 1211 |
+
"grad_norm": 4.624776363372803,
|
| 1212 |
+
"learning_rate": 2.552529305823249e-05,
|
| 1213 |
+
"loss": 4.3225,
|
| 1214 |
+
"step": 15500
|
| 1215 |
+
},
|
| 1216 |
+
{
|
| 1217 |
+
"epoch": 0.4928972246094737,
|
| 1218 |
+
"grad_norm": 4.362007141113281,
|
| 1219 |
+
"learning_rate": 2.5367310183576103e-05,
|
| 1220 |
+
"loss": 4.3164,
|
| 1221 |
+
"step": 15600
|
| 1222 |
+
},
|
| 1223 |
+
{
|
| 1224 |
+
"epoch": 0.4960568222031242,
|
| 1225 |
+
"grad_norm": 3.67773175239563,
|
| 1226 |
+
"learning_rate": 2.5209327308919716e-05,
|
| 1227 |
+
"loss": 4.2915,
|
| 1228 |
+
"step": 15700
|
| 1229 |
+
},
|
| 1230 |
+
{
|
| 1231 |
+
"epoch": 0.49921641979677467,
|
| 1232 |
+
"grad_norm": 3.7847766876220703,
|
| 1233 |
+
"learning_rate": 2.505134443426333e-05,
|
| 1234 |
+
"loss": 4.2696,
|
| 1235 |
+
"step": 15800
|
| 1236 |
+
},
|
| 1237 |
+
{
|
| 1238 |
+
"epoch": 0.5023760173904251,
|
| 1239 |
+
"grad_norm": 4.13474178314209,
|
| 1240 |
+
"learning_rate": 2.489336155960694e-05,
|
| 1241 |
+
"loss": 4.3031,
|
| 1242 |
+
"step": 15900
|
| 1243 |
+
},
|
| 1244 |
+
{
|
| 1245 |
+
"epoch": 0.5055356149840756,
|
| 1246 |
+
"grad_norm": 4.143453598022461,
|
| 1247 |
+
"learning_rate": 2.4735378684950554e-05,
|
| 1248 |
+
"loss": 4.3384,
|
| 1249 |
+
"step": 16000
|
| 1250 |
+
},
|
| 1251 |
+
{
|
| 1252 |
+
"epoch": 0.5055356149840756,
|
| 1253 |
+
"eval_loss": 4.452681541442871,
|
| 1254 |
+
"eval_runtime": 706.2531,
|
| 1255 |
+
"eval_samples_per_second": 124.482,
|
| 1256 |
+
"eval_steps_per_second": 15.561,
|
| 1257 |
+
"step": 16000
|
| 1258 |
+
},
|
| 1259 |
+
{
|
| 1260 |
+
"epoch": 0.5086952125777261,
|
| 1261 |
+
"grad_norm": 4.447348594665527,
|
| 1262 |
+
"learning_rate": 2.4577395810294166e-05,
|
| 1263 |
+
"loss": 4.2958,
|
| 1264 |
+
"step": 16100
|
| 1265 |
+
},
|
| 1266 |
+
{
|
| 1267 |
+
"epoch": 0.5118548101713766,
|
| 1268 |
+
"grad_norm": 4.753751754760742,
|
| 1269 |
+
"learning_rate": 2.441941293563778e-05,
|
| 1270 |
+
"loss": 4.3129,
|
| 1271 |
+
"step": 16200
|
| 1272 |
+
},
|
| 1273 |
+
{
|
| 1274 |
+
"epoch": 0.515014407765027,
|
| 1275 |
+
"grad_norm": 3.83662748336792,
|
| 1276 |
+
"learning_rate": 2.426143006098139e-05,
|
| 1277 |
+
"loss": 4.3168,
|
| 1278 |
+
"step": 16300
|
| 1279 |
+
},
|
| 1280 |
+
{
|
| 1281 |
+
"epoch": 0.5181740053586775,
|
| 1282 |
+
"grad_norm": 3.786252737045288,
|
| 1283 |
+
"learning_rate": 2.4103447186325004e-05,
|
| 1284 |
+
"loss": 4.2844,
|
| 1285 |
+
"step": 16400
|
| 1286 |
+
},
|
| 1287 |
+
{
|
| 1288 |
+
"epoch": 0.521333602952328,
|
| 1289 |
+
"grad_norm": 3.9351420402526855,
|
| 1290 |
+
"learning_rate": 2.3945464311668617e-05,
|
| 1291 |
+
"loss": 4.2816,
|
| 1292 |
+
"step": 16500
|
| 1293 |
+
},
|
| 1294 |
+
{
|
| 1295 |
+
"epoch": 0.5244932005459785,
|
| 1296 |
+
"grad_norm": 4.529867172241211,
|
| 1297 |
+
"learning_rate": 2.378748143701223e-05,
|
| 1298 |
+
"loss": 4.302,
|
| 1299 |
+
"step": 16600
|
| 1300 |
+
},
|
| 1301 |
+
{
|
| 1302 |
+
"epoch": 0.527652798139629,
|
| 1303 |
+
"grad_norm": 4.681921005249023,
|
| 1304 |
+
"learning_rate": 2.3629498562355842e-05,
|
| 1305 |
+
"loss": 4.2904,
|
| 1306 |
+
"step": 16700
|
| 1307 |
+
},
|
| 1308 |
+
{
|
| 1309 |
+
"epoch": 0.5308123957332794,
|
| 1310 |
+
"grad_norm": 3.3952362537384033,
|
| 1311 |
+
"learning_rate": 2.3471515687699455e-05,
|
| 1312 |
+
"loss": 4.2876,
|
| 1313 |
+
"step": 16800
|
| 1314 |
+
},
|
| 1315 |
+
{
|
| 1316 |
+
"epoch": 0.5339719933269299,
|
| 1317 |
+
"grad_norm": 4.524105548858643,
|
| 1318 |
+
"learning_rate": 2.3313532813043067e-05,
|
| 1319 |
+
"loss": 4.262,
|
| 1320 |
+
"step": 16900
|
| 1321 |
+
},
|
| 1322 |
+
{
|
| 1323 |
+
"epoch": 0.5371315909205804,
|
| 1324 |
+
"grad_norm": 4.248294830322266,
|
| 1325 |
+
"learning_rate": 2.315554993838668e-05,
|
| 1326 |
+
"loss": 4.2975,
|
| 1327 |
+
"step": 17000
|
| 1328 |
+
},
|
| 1329 |
+
{
|
| 1330 |
+
"epoch": 0.5371315909205804,
|
| 1331 |
+
"eval_loss": 4.445527076721191,
|
| 1332 |
+
"eval_runtime": 705.3764,
|
| 1333 |
+
"eval_samples_per_second": 124.637,
|
| 1334 |
+
"eval_steps_per_second": 15.58,
|
| 1335 |
+
"step": 17000
|
| 1336 |
+
},
|
| 1337 |
+
{
|
| 1338 |
+
"epoch": 0.5402911885142309,
|
| 1339 |
+
"grad_norm": 4.230813026428223,
|
| 1340 |
+
"learning_rate": 2.2997567063730293e-05,
|
| 1341 |
+
"loss": 4.2806,
|
| 1342 |
+
"step": 17100
|
| 1343 |
+
},
|
| 1344 |
+
{
|
| 1345 |
+
"epoch": 0.5434507861078813,
|
| 1346 |
+
"grad_norm": 3.7945194244384766,
|
| 1347 |
+
"learning_rate": 2.2839584189073905e-05,
|
| 1348 |
+
"loss": 4.2942,
|
| 1349 |
+
"step": 17200
|
| 1350 |
+
},
|
| 1351 |
+
{
|
| 1352 |
+
"epoch": 0.5466103837015318,
|
| 1353 |
+
"grad_norm": 4.855790138244629,
|
| 1354 |
+
"learning_rate": 2.2681601314417518e-05,
|
| 1355 |
+
"loss": 4.2818,
|
| 1356 |
+
"step": 17300
|
| 1357 |
+
},
|
| 1358 |
+
{
|
| 1359 |
+
"epoch": 0.5497699812951823,
|
| 1360 |
+
"grad_norm": 4.7962236404418945,
|
| 1361 |
+
"learning_rate": 2.252361843976113e-05,
|
| 1362 |
+
"loss": 4.2804,
|
| 1363 |
+
"step": 17400
|
| 1364 |
+
},
|
| 1365 |
+
{
|
| 1366 |
+
"epoch": 0.5529295788888328,
|
| 1367 |
+
"grad_norm": 4.337451934814453,
|
| 1368 |
+
"learning_rate": 2.2365635565104747e-05,
|
| 1369 |
+
"loss": 4.3063,
|
| 1370 |
+
"step": 17500
|
| 1371 |
+
},
|
| 1372 |
+
{
|
| 1373 |
+
"epoch": 0.5560891764824832,
|
| 1374 |
+
"grad_norm": 4.422916889190674,
|
| 1375 |
+
"learning_rate": 2.220765269044836e-05,
|
| 1376 |
+
"loss": 4.2942,
|
| 1377 |
+
"step": 17600
|
| 1378 |
+
},
|
| 1379 |
+
{
|
| 1380 |
+
"epoch": 0.5592487740761337,
|
| 1381 |
+
"grad_norm": 4.27534294128418,
|
| 1382 |
+
"learning_rate": 2.204966981579197e-05,
|
| 1383 |
+
"loss": 4.2828,
|
| 1384 |
+
"step": 17700
|
| 1385 |
+
},
|
| 1386 |
+
{
|
| 1387 |
+
"epoch": 0.5624083716697842,
|
| 1388 |
+
"grad_norm": 4.3357648849487305,
|
| 1389 |
+
"learning_rate": 2.1893266769882147e-05,
|
| 1390 |
+
"loss": 4.2766,
|
| 1391 |
+
"step": 17800
|
| 1392 |
+
},
|
| 1393 |
+
{
|
| 1394 |
+
"epoch": 0.5655679692634346,
|
| 1395 |
+
"grad_norm": 3.859369993209839,
|
| 1396 |
+
"learning_rate": 2.173528389522576e-05,
|
| 1397 |
+
"loss": 4.2979,
|
| 1398 |
+
"step": 17900
|
| 1399 |
+
},
|
| 1400 |
+
{
|
| 1401 |
+
"epoch": 0.568727566857085,
|
| 1402 |
+
"grad_norm": 5.381210803985596,
|
| 1403 |
+
"learning_rate": 2.1577301020569372e-05,
|
| 1404 |
+
"loss": 4.3169,
|
| 1405 |
+
"step": 18000
|
| 1406 |
+
},
|
| 1407 |
+
{
|
| 1408 |
+
"epoch": 0.568727566857085,
|
| 1409 |
+
"eval_loss": 4.439344882965088,
|
| 1410 |
+
"eval_runtime": 708.2829,
|
| 1411 |
+
"eval_samples_per_second": 124.126,
|
| 1412 |
+
"eval_steps_per_second": 15.516,
|
| 1413 |
+
"step": 18000
|
| 1414 |
+
},
|
| 1415 |
+
{
|
| 1416 |
+
"epoch": 0.5718871644507355,
|
| 1417 |
+
"grad_norm": 3.3209519386291504,
|
| 1418 |
+
"learning_rate": 2.1419318145912985e-05,
|
| 1419 |
+
"loss": 4.2588,
|
| 1420 |
+
"step": 18100
|
| 1421 |
+
},
|
| 1422 |
+
{
|
| 1423 |
+
"epoch": 0.575046762044386,
|
| 1424 |
+
"grad_norm": 3.9373040199279785,
|
| 1425 |
+
"learning_rate": 2.1261335271256597e-05,
|
| 1426 |
+
"loss": 4.2895,
|
| 1427 |
+
"step": 18200
|
| 1428 |
+
},
|
| 1429 |
+
{
|
| 1430 |
+
"epoch": 0.5782063596380365,
|
| 1431 |
+
"grad_norm": 3.453148126602173,
|
| 1432 |
+
"learning_rate": 2.110335239660021e-05,
|
| 1433 |
+
"loss": 4.276,
|
| 1434 |
+
"step": 18300
|
| 1435 |
+
},
|
| 1436 |
+
{
|
| 1437 |
+
"epoch": 0.5813659572316869,
|
| 1438 |
+
"grad_norm": 3.7350664138793945,
|
| 1439 |
+
"learning_rate": 2.0945369521943823e-05,
|
| 1440 |
+
"loss": 4.3268,
|
| 1441 |
+
"step": 18400
|
| 1442 |
+
},
|
| 1443 |
+
{
|
| 1444 |
+
"epoch": 0.5845255548253374,
|
| 1445 |
+
"grad_norm": 4.277946949005127,
|
| 1446 |
+
"learning_rate": 2.0787386647287435e-05,
|
| 1447 |
+
"loss": 4.2573,
|
| 1448 |
+
"step": 18500
|
| 1449 |
+
},
|
| 1450 |
+
{
|
| 1451 |
+
"epoch": 0.5876851524189879,
|
| 1452 |
+
"grad_norm": 4.000477313995361,
|
| 1453 |
+
"learning_rate": 2.0629403772631048e-05,
|
| 1454 |
+
"loss": 4.2568,
|
| 1455 |
+
"step": 18600
|
| 1456 |
+
},
|
| 1457 |
+
{
|
| 1458 |
+
"epoch": 0.5908447500126384,
|
| 1459 |
+
"grad_norm": 3.6891930103302,
|
| 1460 |
+
"learning_rate": 2.047142089797466e-05,
|
| 1461 |
+
"loss": 4.2784,
|
| 1462 |
+
"step": 18700
|
| 1463 |
+
},
|
| 1464 |
+
{
|
| 1465 |
+
"epoch": 0.5940043476062888,
|
| 1466 |
+
"grad_norm": 4.800564765930176,
|
| 1467 |
+
"learning_rate": 2.0313438023318273e-05,
|
| 1468 |
+
"loss": 4.3016,
|
| 1469 |
+
"step": 18800
|
| 1470 |
+
},
|
| 1471 |
+
{
|
| 1472 |
+
"epoch": 0.5971639451999393,
|
| 1473 |
+
"grad_norm": 5.003584384918213,
|
| 1474 |
+
"learning_rate": 2.0155455148661886e-05,
|
| 1475 |
+
"loss": 4.2818,
|
| 1476 |
+
"step": 18900
|
| 1477 |
+
},
|
| 1478 |
+
{
|
| 1479 |
+
"epoch": 0.6003235427935898,
|
| 1480 |
+
"grad_norm": 4.17577600479126,
|
| 1481 |
+
"learning_rate": 1.99974722740055e-05,
|
| 1482 |
+
"loss": 4.2794,
|
| 1483 |
+
"step": 19000
|
| 1484 |
+
},
|
| 1485 |
+
{
|
| 1486 |
+
"epoch": 0.6003235427935898,
|
| 1487 |
+
"eval_loss": 4.433346748352051,
|
| 1488 |
+
"eval_runtime": 712.1597,
|
| 1489 |
+
"eval_samples_per_second": 123.45,
|
| 1490 |
+
"eval_steps_per_second": 15.432,
|
| 1491 |
+
"step": 19000
|
| 1492 |
+
},
|
| 1493 |
+
{
|
| 1494 |
+
"epoch": 0.6034831403872403,
|
| 1495 |
+
"grad_norm": 4.12880277633667,
|
| 1496 |
+
"learning_rate": 1.983948939934911e-05,
|
| 1497 |
+
"loss": 4.2539,
|
| 1498 |
+
"step": 19100
|
| 1499 |
+
},
|
| 1500 |
+
{
|
| 1501 |
+
"epoch": 0.6066427379808907,
|
| 1502 |
+
"grad_norm": 4.599102020263672,
|
| 1503 |
+
"learning_rate": 1.9681506524692724e-05,
|
| 1504 |
+
"loss": 4.2842,
|
| 1505 |
+
"step": 19200
|
| 1506 |
+
},
|
| 1507 |
+
{
|
| 1508 |
+
"epoch": 0.6098023355745412,
|
| 1509 |
+
"grad_norm": 4.2362380027771,
|
| 1510 |
+
"learning_rate": 1.9523523650036336e-05,
|
| 1511 |
+
"loss": 4.2803,
|
| 1512 |
+
"step": 19300
|
| 1513 |
+
},
|
| 1514 |
+
{
|
| 1515 |
+
"epoch": 0.6129619331681917,
|
| 1516 |
+
"grad_norm": 4.119621753692627,
|
| 1517 |
+
"learning_rate": 1.936554077537995e-05,
|
| 1518 |
+
"loss": 4.2413,
|
| 1519 |
+
"step": 19400
|
| 1520 |
+
},
|
| 1521 |
+
{
|
| 1522 |
+
"epoch": 0.6161215307618422,
|
| 1523 |
+
"grad_norm": 4.374582290649414,
|
| 1524 |
+
"learning_rate": 1.9207557900723565e-05,
|
| 1525 |
+
"loss": 4.3056,
|
| 1526 |
+
"step": 19500
|
| 1527 |
+
},
|
| 1528 |
+
{
|
| 1529 |
+
"epoch": 0.6192811283554926,
|
| 1530 |
+
"grad_norm": 3.5877952575683594,
|
| 1531 |
+
"learning_rate": 1.9049575026067178e-05,
|
| 1532 |
+
"loss": 4.2978,
|
| 1533 |
+
"step": 19600
|
| 1534 |
+
},
|
| 1535 |
+
{
|
| 1536 |
+
"epoch": 0.6224407259491431,
|
| 1537 |
+
"grad_norm": 4.062015056610107,
|
| 1538 |
+
"learning_rate": 1.889159215141079e-05,
|
| 1539 |
+
"loss": 4.2525,
|
| 1540 |
+
"step": 19700
|
| 1541 |
+
},
|
| 1542 |
+
{
|
| 1543 |
+
"epoch": 0.6256003235427936,
|
| 1544 |
+
"grad_norm": 4.872693061828613,
|
| 1545 |
+
"learning_rate": 1.8733609276754403e-05,
|
| 1546 |
+
"loss": 4.3039,
|
| 1547 |
+
"step": 19800
|
| 1548 |
+
},
|
| 1549 |
+
{
|
| 1550 |
+
"epoch": 0.6287599211364441,
|
| 1551 |
+
"grad_norm": 3.899491310119629,
|
| 1552 |
+
"learning_rate": 1.8575626402098016e-05,
|
| 1553 |
+
"loss": 4.3016,
|
| 1554 |
+
"step": 19900
|
| 1555 |
+
},
|
| 1556 |
+
{
|
| 1557 |
+
"epoch": 0.6319195187300946,
|
| 1558 |
+
"grad_norm": 3.927971601486206,
|
| 1559 |
+
"learning_rate": 1.8417643527441628e-05,
|
| 1560 |
+
"loss": 4.2883,
|
| 1561 |
+
"step": 20000
|
| 1562 |
+
},
|
| 1563 |
+
{
|
| 1564 |
+
"epoch": 0.6319195187300946,
|
| 1565 |
+
"eval_loss": 4.4285101890563965,
|
| 1566 |
+
"eval_runtime": 708.5523,
|
| 1567 |
+
"eval_samples_per_second": 124.078,
|
| 1568 |
+
"eval_steps_per_second": 15.51,
|
| 1569 |
+
"step": 20000
|
| 1570 |
+
},
|
| 1571 |
+
{
|
| 1572 |
+
"epoch": 0.635079116323745,
|
| 1573 |
+
"grad_norm": 3.7342617511749268,
|
| 1574 |
+
"learning_rate": 1.8259660652785237e-05,
|
| 1575 |
+
"loss": 4.2633,
|
| 1576 |
+
"step": 20100
|
| 1577 |
+
},
|
| 1578 |
+
{
|
| 1579 |
+
"epoch": 0.6382387139173955,
|
| 1580 |
+
"grad_norm": 5.151529788970947,
|
| 1581 |
+
"learning_rate": 1.810167777812885e-05,
|
| 1582 |
+
"loss": 4.2858,
|
| 1583 |
+
"step": 20200
|
| 1584 |
+
},
|
| 1585 |
+
{
|
| 1586 |
+
"epoch": 0.641398311511046,
|
| 1587 |
+
"grad_norm": 6.981259822845459,
|
| 1588 |
+
"learning_rate": 1.7943694903472463e-05,
|
| 1589 |
+
"loss": 4.2812,
|
| 1590 |
+
"step": 20300
|
| 1591 |
+
},
|
| 1592 |
+
{
|
| 1593 |
+
"epoch": 0.6445579091046965,
|
| 1594 |
+
"grad_norm": 4.146710395812988,
|
| 1595 |
+
"learning_rate": 1.7785712028816075e-05,
|
| 1596 |
+
"loss": 4.295,
|
| 1597 |
+
"step": 20400
|
| 1598 |
+
},
|
| 1599 |
+
{
|
| 1600 |
+
"epoch": 0.6477175066983469,
|
| 1601 |
+
"grad_norm": 3.6891162395477295,
|
| 1602 |
+
"learning_rate": 1.7629308982906254e-05,
|
| 1603 |
+
"loss": 4.268,
|
| 1604 |
+
"step": 20500
|
| 1605 |
+
},
|
| 1606 |
+
{
|
| 1607 |
+
"epoch": 0.6508771042919974,
|
| 1608 |
+
"grad_norm": 4.074377536773682,
|
| 1609 |
+
"learning_rate": 1.7471326108249866e-05,
|
| 1610 |
+
"loss": 4.2796,
|
| 1611 |
+
"step": 20600
|
| 1612 |
+
},
|
| 1613 |
+
{
|
| 1614 |
+
"epoch": 0.6540367018856479,
|
| 1615 |
+
"grad_norm": 3.6414151191711426,
|
| 1616 |
+
"learning_rate": 1.731334323359348e-05,
|
| 1617 |
+
"loss": 4.2815,
|
| 1618 |
+
"step": 20700
|
| 1619 |
+
},
|
| 1620 |
+
{
|
| 1621 |
+
"epoch": 0.6571962994792984,
|
| 1622 |
+
"grad_norm": 4.212449073791504,
|
| 1623 |
+
"learning_rate": 1.7156940187683657e-05,
|
| 1624 |
+
"loss": 4.2843,
|
| 1625 |
+
"step": 20800
|
| 1626 |
+
},
|
| 1627 |
+
{
|
| 1628 |
+
"epoch": 0.6603558970729488,
|
| 1629 |
+
"grad_norm": 4.951251983642578,
|
| 1630 |
+
"learning_rate": 1.699895731302727e-05,
|
| 1631 |
+
"loss": 4.271,
|
| 1632 |
+
"step": 20900
|
| 1633 |
+
},
|
| 1634 |
+
{
|
| 1635 |
+
"epoch": 0.6635154946665993,
|
| 1636 |
+
"grad_norm": 5.257840633392334,
|
| 1637 |
+
"learning_rate": 1.6840974438370883e-05,
|
| 1638 |
+
"loss": 4.2744,
|
| 1639 |
+
"step": 21000
|
| 1640 |
+
},
|
| 1641 |
+
{
|
| 1642 |
+
"epoch": 0.6635154946665993,
|
| 1643 |
+
"eval_loss": 4.42351770401001,
|
| 1644 |
+
"eval_runtime": 703.9002,
|
| 1645 |
+
"eval_samples_per_second": 124.898,
|
| 1646 |
+
"eval_steps_per_second": 15.613,
|
| 1647 |
+
"step": 21000
|
| 1648 |
+
},
|
| 1649 |
+
{
|
| 1650 |
+
"epoch": 0.6666750922602497,
|
| 1651 |
+
"grad_norm": 4.769164562225342,
|
| 1652 |
+
"learning_rate": 1.6682991563714495e-05,
|
| 1653 |
+
"loss": 4.2332,
|
| 1654 |
+
"step": 21100
|
| 1655 |
+
},
|
| 1656 |
+
{
|
| 1657 |
+
"epoch": 0.6698346898539002,
|
| 1658 |
+
"grad_norm": 3.9750113487243652,
|
| 1659 |
+
"learning_rate": 1.6525008689058108e-05,
|
| 1660 |
+
"loss": 4.256,
|
| 1661 |
+
"step": 21200
|
| 1662 |
+
},
|
| 1663 |
+
{
|
| 1664 |
+
"epoch": 0.6729942874475506,
|
| 1665 |
+
"grad_norm": 4.802659511566162,
|
| 1666 |
+
"learning_rate": 1.636702581440172e-05,
|
| 1667 |
+
"loss": 4.2465,
|
| 1668 |
+
"step": 21300
|
| 1669 |
+
},
|
| 1670 |
+
{
|
| 1671 |
+
"epoch": 0.6761538850412011,
|
| 1672 |
+
"grad_norm": 4.48084831237793,
|
| 1673 |
+
"learning_rate": 1.6209042939745333e-05,
|
| 1674 |
+
"loss": 4.2762,
|
| 1675 |
+
"step": 21400
|
| 1676 |
+
},
|
| 1677 |
+
{
|
| 1678 |
+
"epoch": 0.6793134826348516,
|
| 1679 |
+
"grad_norm": 5.000456809997559,
|
| 1680 |
+
"learning_rate": 1.6051060065088946e-05,
|
| 1681 |
+
"loss": 4.2937,
|
| 1682 |
+
"step": 21500
|
| 1683 |
+
},
|
| 1684 |
+
{
|
| 1685 |
+
"epoch": 0.6824730802285021,
|
| 1686 |
+
"grad_norm": 4.686836242675781,
|
| 1687 |
+
"learning_rate": 1.5893077190432555e-05,
|
| 1688 |
+
"loss": 4.2684,
|
| 1689 |
+
"step": 21600
|
| 1690 |
+
},
|
| 1691 |
+
{
|
| 1692 |
+
"epoch": 0.6856326778221525,
|
| 1693 |
+
"grad_norm": 3.528893232345581,
|
| 1694 |
+
"learning_rate": 1.5735094315776168e-05,
|
| 1695 |
+
"loss": 4.2491,
|
| 1696 |
+
"step": 21700
|
| 1697 |
+
},
|
| 1698 |
+
{
|
| 1699 |
+
"epoch": 0.688792275415803,
|
| 1700 |
+
"grad_norm": 4.837214946746826,
|
| 1701 |
+
"learning_rate": 1.5577111441119784e-05,
|
| 1702 |
+
"loss": 4.2601,
|
| 1703 |
+
"step": 21800
|
| 1704 |
+
},
|
| 1705 |
+
{
|
| 1706 |
+
"epoch": 0.6919518730094535,
|
| 1707 |
+
"grad_norm": 4.445570468902588,
|
| 1708 |
+
"learning_rate": 1.5419128566463396e-05,
|
| 1709 |
+
"loss": 4.2533,
|
| 1710 |
+
"step": 21900
|
| 1711 |
+
},
|
| 1712 |
+
{
|
| 1713 |
+
"epoch": 0.695111470603104,
|
| 1714 |
+
"grad_norm": 4.073211193084717,
|
| 1715 |
+
"learning_rate": 1.526114569180701e-05,
|
| 1716 |
+
"loss": 4.2732,
|
| 1717 |
+
"step": 22000
|
| 1718 |
+
},
|
| 1719 |
+
{
|
| 1720 |
+
"epoch": 0.695111470603104,
|
| 1721 |
+
"eval_loss": 4.418666839599609,
|
| 1722 |
+
"eval_runtime": 706.2044,
|
| 1723 |
+
"eval_samples_per_second": 124.491,
|
| 1724 |
+
"eval_steps_per_second": 15.562,
|
| 1725 |
+
"step": 22000
|
| 1726 |
+
},
|
| 1727 |
+
{
|
| 1728 |
+
"epoch": 0.6982710681967544,
|
| 1729 |
+
"grad_norm": 4.091275691986084,
|
| 1730 |
+
"learning_rate": 1.510316281715062e-05,
|
| 1731 |
+
"loss": 4.2226,
|
| 1732 |
+
"step": 22100
|
| 1733 |
+
},
|
| 1734 |
+
{
|
| 1735 |
+
"epoch": 0.7014306657904049,
|
| 1736 |
+
"grad_norm": 4.8102192878723145,
|
| 1737 |
+
"learning_rate": 1.4945179942494233e-05,
|
| 1738 |
+
"loss": 4.2794,
|
| 1739 |
+
"step": 22200
|
| 1740 |
+
},
|
| 1741 |
+
{
|
| 1742 |
+
"epoch": 0.7045902633840554,
|
| 1743 |
+
"grad_norm": 4.450103282928467,
|
| 1744 |
+
"learning_rate": 1.4787197067837847e-05,
|
| 1745 |
+
"loss": 4.2869,
|
| 1746 |
+
"step": 22300
|
| 1747 |
+
},
|
| 1748 |
+
{
|
| 1749 |
+
"epoch": 0.7077498609777059,
|
| 1750 |
+
"grad_norm": 4.858384609222412,
|
| 1751 |
+
"learning_rate": 1.462921419318146e-05,
|
| 1752 |
+
"loss": 4.2616,
|
| 1753 |
+
"step": 22400
|
| 1754 |
+
},
|
| 1755 |
+
{
|
| 1756 |
+
"epoch": 0.7109094585713563,
|
| 1757 |
+
"grad_norm": 4.810959339141846,
|
| 1758 |
+
"learning_rate": 1.4471231318525072e-05,
|
| 1759 |
+
"loss": 4.2589,
|
| 1760 |
+
"step": 22500
|
| 1761 |
+
},
|
| 1762 |
+
{
|
| 1763 |
+
"epoch": 0.7140690561650068,
|
| 1764 |
+
"grad_norm": 4.832188129425049,
|
| 1765 |
+
"learning_rate": 1.4313248443868685e-05,
|
| 1766 |
+
"loss": 4.2668,
|
| 1767 |
+
"step": 22600
|
| 1768 |
+
},
|
| 1769 |
+
{
|
| 1770 |
+
"epoch": 0.7172286537586573,
|
| 1771 |
+
"grad_norm": 4.8991475105285645,
|
| 1772 |
+
"learning_rate": 1.4155265569212298e-05,
|
| 1773 |
+
"loss": 4.2759,
|
| 1774 |
+
"step": 22700
|
| 1775 |
+
},
|
| 1776 |
+
{
|
| 1777 |
+
"epoch": 0.7203882513523078,
|
| 1778 |
+
"grad_norm": 5.653695106506348,
|
| 1779 |
+
"learning_rate": 1.399728269455591e-05,
|
| 1780 |
+
"loss": 4.2745,
|
| 1781 |
+
"step": 22800
|
| 1782 |
+
},
|
| 1783 |
+
{
|
| 1784 |
+
"epoch": 0.7235478489459582,
|
| 1785 |
+
"grad_norm": 4.475470542907715,
|
| 1786 |
+
"learning_rate": 1.3839299819899523e-05,
|
| 1787 |
+
"loss": 4.2785,
|
| 1788 |
+
"step": 22900
|
| 1789 |
+
},
|
| 1790 |
+
{
|
| 1791 |
+
"epoch": 0.7267074465396087,
|
| 1792 |
+
"grad_norm": 4.225826263427734,
|
| 1793 |
+
"learning_rate": 1.3681316945243135e-05,
|
| 1794 |
+
"loss": 4.257,
|
| 1795 |
+
"step": 23000
|
| 1796 |
+
},
|
| 1797 |
+
{
|
| 1798 |
+
"epoch": 0.7267074465396087,
|
| 1799 |
+
"eval_loss": 4.414594650268555,
|
| 1800 |
+
"eval_runtime": 709.538,
|
| 1801 |
+
"eval_samples_per_second": 123.906,
|
| 1802 |
+
"eval_steps_per_second": 15.489,
|
| 1803 |
+
"step": 23000
|
| 1804 |
+
},
|
| 1805 |
+
{
|
| 1806 |
+
"epoch": 0.7298670441332592,
|
| 1807 |
+
"grad_norm": 5.017648220062256,
|
| 1808 |
+
"learning_rate": 1.352333407058675e-05,
|
| 1809 |
+
"loss": 4.3036,
|
| 1810 |
+
"step": 23100
|
| 1811 |
+
},
|
| 1812 |
+
{
|
| 1813 |
+
"epoch": 0.7330266417269097,
|
| 1814 |
+
"grad_norm": 4.95635986328125,
|
| 1815 |
+
"learning_rate": 1.3365351195930362e-05,
|
| 1816 |
+
"loss": 4.295,
|
| 1817 |
+
"step": 23200
|
| 1818 |
+
},
|
| 1819 |
+
{
|
| 1820 |
+
"epoch": 0.7361862393205602,
|
| 1821 |
+
"grad_norm": 5.106438159942627,
|
| 1822 |
+
"learning_rate": 1.3207368321273975e-05,
|
| 1823 |
+
"loss": 4.2452,
|
| 1824 |
+
"step": 23300
|
| 1825 |
+
},
|
| 1826 |
+
{
|
| 1827 |
+
"epoch": 0.7393458369142106,
|
| 1828 |
+
"grad_norm": 5.10938024520874,
|
| 1829 |
+
"learning_rate": 1.3049385446617588e-05,
|
| 1830 |
+
"loss": 4.2532,
|
| 1831 |
+
"step": 23400
|
| 1832 |
+
},
|
| 1833 |
+
{
|
| 1834 |
+
"epoch": 0.7425054345078611,
|
| 1835 |
+
"grad_norm": 4.140163421630859,
|
| 1836 |
+
"learning_rate": 1.28914025719612e-05,
|
| 1837 |
+
"loss": 4.262,
|
| 1838 |
+
"step": 23500
|
| 1839 |
+
},
|
| 1840 |
+
{
|
| 1841 |
+
"epoch": 0.7456650321015116,
|
| 1842 |
+
"grad_norm": 4.146867275238037,
|
| 1843 |
+
"learning_rate": 1.2733419697304813e-05,
|
| 1844 |
+
"loss": 4.2381,
|
| 1845 |
+
"step": 23600
|
| 1846 |
+
},
|
| 1847 |
+
{
|
| 1848 |
+
"epoch": 0.7488246296951621,
|
| 1849 |
+
"grad_norm": 4.620384216308594,
|
| 1850 |
+
"learning_rate": 1.2575436822648426e-05,
|
| 1851 |
+
"loss": 4.2509,
|
| 1852 |
+
"step": 23700
|
| 1853 |
+
},
|
| 1854 |
+
{
|
| 1855 |
+
"epoch": 0.7519842272888125,
|
| 1856 |
+
"grad_norm": 4.155729293823242,
|
| 1857 |
+
"learning_rate": 1.2417453947992038e-05,
|
| 1858 |
+
"loss": 4.2816,
|
| 1859 |
+
"step": 23800
|
| 1860 |
+
},
|
| 1861 |
+
{
|
| 1862 |
+
"epoch": 0.755143824882463,
|
| 1863 |
+
"grad_norm": 6.052363395690918,
|
| 1864 |
+
"learning_rate": 1.2259471073335651e-05,
|
| 1865 |
+
"loss": 4.2619,
|
| 1866 |
+
"step": 23900
|
| 1867 |
+
},
|
| 1868 |
+
{
|
| 1869 |
+
"epoch": 0.7583034224761135,
|
| 1870 |
+
"grad_norm": 4.320205211639404,
|
| 1871 |
+
"learning_rate": 1.2101488198679265e-05,
|
| 1872 |
+
"loss": 4.2691,
|
| 1873 |
+
"step": 24000
|
| 1874 |
+
},
|
| 1875 |
+
{
|
| 1876 |
+
"epoch": 0.7583034224761135,
|
| 1877 |
+
"eval_loss": 4.410797595977783,
|
| 1878 |
+
"eval_runtime": 708.2786,
|
| 1879 |
+
"eval_samples_per_second": 124.126,
|
| 1880 |
+
"eval_steps_per_second": 15.516,
|
| 1881 |
+
"step": 24000
|
| 1882 |
+
},
|
| 1883 |
+
{
|
| 1884 |
+
"epoch": 0.761463020069764,
|
| 1885 |
+
"grad_norm": 4.142159461975098,
|
| 1886 |
+
"learning_rate": 1.1943505324022876e-05,
|
| 1887 |
+
"loss": 4.277,
|
| 1888 |
+
"step": 24100
|
| 1889 |
+
},
|
| 1890 |
+
{
|
| 1891 |
+
"epoch": 0.7646226176634144,
|
| 1892 |
+
"grad_norm": 4.465941905975342,
|
| 1893 |
+
"learning_rate": 1.1785522449366489e-05,
|
| 1894 |
+
"loss": 4.2629,
|
| 1895 |
+
"step": 24200
|
| 1896 |
+
},
|
| 1897 |
+
{
|
| 1898 |
+
"epoch": 0.7677822152570649,
|
| 1899 |
+
"grad_norm": 3.3799822330474854,
|
| 1900 |
+
"learning_rate": 1.1627539574710101e-05,
|
| 1901 |
+
"loss": 4.2882,
|
| 1902 |
+
"step": 24300
|
| 1903 |
+
},
|
| 1904 |
+
{
|
| 1905 |
+
"epoch": 0.7709418128507153,
|
| 1906 |
+
"grad_norm": 4.531765460968018,
|
| 1907 |
+
"learning_rate": 1.1469556700053714e-05,
|
| 1908 |
+
"loss": 4.2546,
|
| 1909 |
+
"step": 24400
|
| 1910 |
+
},
|
| 1911 |
+
{
|
| 1912 |
+
"epoch": 0.7741014104443658,
|
| 1913 |
+
"grad_norm": 4.775002956390381,
|
| 1914 |
+
"learning_rate": 1.1311573825397327e-05,
|
| 1915 |
+
"loss": 4.2376,
|
| 1916 |
+
"step": 24500
|
| 1917 |
+
},
|
| 1918 |
+
{
|
| 1919 |
+
"epoch": 0.7772610080380162,
|
| 1920 |
+
"grad_norm": 5.156617164611816,
|
| 1921 |
+
"learning_rate": 1.115359095074094e-05,
|
| 1922 |
+
"loss": 4.2341,
|
| 1923 |
+
"step": 24600
|
| 1924 |
+
},
|
| 1925 |
+
{
|
| 1926 |
+
"epoch": 0.7804206056316667,
|
| 1927 |
+
"grad_norm": 4.456455230712891,
|
| 1928 |
+
"learning_rate": 1.0995608076084554e-05,
|
| 1929 |
+
"loss": 4.2659,
|
| 1930 |
+
"step": 24700
|
| 1931 |
+
},
|
| 1932 |
+
{
|
| 1933 |
+
"epoch": 0.7835802032253172,
|
| 1934 |
+
"grad_norm": 4.179417133331299,
|
| 1935 |
+
"learning_rate": 1.083920503017473e-05,
|
| 1936 |
+
"loss": 4.2722,
|
| 1937 |
+
"step": 24800
|
| 1938 |
+
},
|
| 1939 |
+
{
|
| 1940 |
+
"epoch": 0.7867398008189677,
|
| 1941 |
+
"grad_norm": 4.3775224685668945,
|
| 1942 |
+
"learning_rate": 1.0681222155518341e-05,
|
| 1943 |
+
"loss": 4.278,
|
| 1944 |
+
"step": 24900
|
| 1945 |
+
},
|
| 1946 |
+
{
|
| 1947 |
+
"epoch": 0.7898993984126181,
|
| 1948 |
+
"grad_norm": 3.773751974105835,
|
| 1949 |
+
"learning_rate": 1.0523239280861954e-05,
|
| 1950 |
+
"loss": 4.2548,
|
| 1951 |
+
"step": 25000
|
| 1952 |
+
},
|
| 1953 |
+
{
|
| 1954 |
+
"epoch": 0.7898993984126181,
|
| 1955 |
+
"eval_loss": 4.407622814178467,
|
| 1956 |
+
"eval_runtime": 708.246,
|
| 1957 |
+
"eval_samples_per_second": 124.132,
|
| 1958 |
+
"eval_steps_per_second": 15.517,
|
| 1959 |
+
"step": 25000
|
| 1960 |
+
},
|
| 1961 |
+
{
|
| 1962 |
+
"epoch": 0.7930589960062686,
|
| 1963 |
+
"grad_norm": 3.457412004470825,
|
| 1964 |
+
"learning_rate": 1.0365256406205567e-05,
|
| 1965 |
+
"loss": 4.2549,
|
| 1966 |
+
"step": 25100
|
| 1967 |
+
},
|
| 1968 |
+
{
|
| 1969 |
+
"epoch": 0.7962185935999191,
|
| 1970 |
+
"grad_norm": 3.731058359146118,
|
| 1971 |
+
"learning_rate": 1.0207273531549181e-05,
|
| 1972 |
+
"loss": 4.2283,
|
| 1973 |
+
"step": 25200
|
| 1974 |
+
},
|
| 1975 |
+
{
|
| 1976 |
+
"epoch": 0.7993781911935696,
|
| 1977 |
+
"grad_norm": 4.568631172180176,
|
| 1978 |
+
"learning_rate": 1.0049290656892794e-05,
|
| 1979 |
+
"loss": 4.2678,
|
| 1980 |
+
"step": 25300
|
| 1981 |
+
},
|
| 1982 |
+
{
|
| 1983 |
+
"epoch": 0.80253778878722,
|
| 1984 |
+
"grad_norm": 4.449248313903809,
|
| 1985 |
+
"learning_rate": 9.891307782236406e-06,
|
| 1986 |
+
"loss": 4.2368,
|
| 1987 |
+
"step": 25400
|
| 1988 |
+
},
|
| 1989 |
+
{
|
| 1990 |
+
"epoch": 0.8056973863808705,
|
| 1991 |
+
"grad_norm": 4.026270389556885,
|
| 1992 |
+
"learning_rate": 9.733324907580019e-06,
|
| 1993 |
+
"loss": 4.2688,
|
| 1994 |
+
"step": 25500
|
| 1995 |
+
},
|
| 1996 |
+
{
|
| 1997 |
+
"epoch": 0.808856983974521,
|
| 1998 |
+
"grad_norm": 4.677060604095459,
|
| 1999 |
+
"learning_rate": 9.575342032923631e-06,
|
| 2000 |
+
"loss": 4.2312,
|
| 2001 |
+
"step": 25600
|
| 2002 |
+
},
|
| 2003 |
+
{
|
| 2004 |
+
"epoch": 0.8120165815681715,
|
| 2005 |
+
"grad_norm": 3.9102447032928467,
|
| 2006 |
+
"learning_rate": 9.417359158267244e-06,
|
| 2007 |
+
"loss": 4.2257,
|
| 2008 |
+
"step": 25700
|
| 2009 |
+
},
|
| 2010 |
+
{
|
| 2011 |
+
"epoch": 0.815176179161822,
|
| 2012 |
+
"grad_norm": 5.438532829284668,
|
| 2013 |
+
"learning_rate": 9.259376283610857e-06,
|
| 2014 |
+
"loss": 4.2285,
|
| 2015 |
+
"step": 25800
|
| 2016 |
+
},
|
| 2017 |
+
{
|
| 2018 |
+
"epoch": 0.8183357767554724,
|
| 2019 |
+
"grad_norm": 4.997466087341309,
|
| 2020 |
+
"learning_rate": 9.10139340895447e-06,
|
| 2021 |
+
"loss": 4.2642,
|
| 2022 |
+
"step": 25900
|
| 2023 |
+
},
|
| 2024 |
+
{
|
| 2025 |
+
"epoch": 0.8214953743491229,
|
| 2026 |
+
"grad_norm": 4.287134647369385,
|
| 2027 |
+
"learning_rate": 8.943410534298082e-06,
|
| 2028 |
+
"loss": 4.2341,
|
| 2029 |
+
"step": 26000
|
| 2030 |
+
},
|
| 2031 |
+
{
|
| 2032 |
+
"epoch": 0.8214953743491229,
|
| 2033 |
+
"eval_loss": 4.404691219329834,
|
| 2034 |
+
"eval_runtime": 706.9434,
|
| 2035 |
+
"eval_samples_per_second": 124.361,
|
| 2036 |
+
"eval_steps_per_second": 15.546,
|
| 2037 |
+
"step": 26000
|
| 2038 |
+
},
|
| 2039 |
+
{
|
| 2040 |
+
"epoch": 0.8246549719427734,
|
| 2041 |
+
"grad_norm": 4.450557708740234,
|
| 2042 |
+
"learning_rate": 8.785427659641696e-06,
|
| 2043 |
+
"loss": 4.289,
|
| 2044 |
+
"step": 26100
|
| 2045 |
+
},
|
| 2046 |
+
{
|
| 2047 |
+
"epoch": 0.8278145695364238,
|
| 2048 |
+
"grad_norm": 3.8313915729522705,
|
| 2049 |
+
"learning_rate": 8.627444784985309e-06,
|
| 2050 |
+
"loss": 4.2494,
|
| 2051 |
+
"step": 26200
|
| 2052 |
+
},
|
| 2053 |
+
{
|
| 2054 |
+
"epoch": 0.8309741671300743,
|
| 2055 |
+
"grad_norm": 4.760807514190674,
|
| 2056 |
+
"learning_rate": 8.469461910328922e-06,
|
| 2057 |
+
"loss": 4.2273,
|
| 2058 |
+
"step": 26300
|
| 2059 |
+
},
|
| 2060 |
+
{
|
| 2061 |
+
"epoch": 0.8341337647237248,
|
| 2062 |
+
"grad_norm": 4.606013774871826,
|
| 2063 |
+
"learning_rate": 8.311479035672534e-06,
|
| 2064 |
+
"loss": 4.2665,
|
| 2065 |
+
"step": 26400
|
| 2066 |
+
},
|
| 2067 |
+
{
|
| 2068 |
+
"epoch": 0.8372933623173753,
|
| 2069 |
+
"grad_norm": 3.884476661682129,
|
| 2070 |
+
"learning_rate": 8.153496161016145e-06,
|
| 2071 |
+
"loss": 4.2543,
|
| 2072 |
+
"step": 26500
|
| 2073 |
+
},
|
| 2074 |
+
{
|
| 2075 |
+
"epoch": 0.8404529599110258,
|
| 2076 |
+
"grad_norm": 4.1962127685546875,
|
| 2077 |
+
"learning_rate": 7.995513286359758e-06,
|
| 2078 |
+
"loss": 4.2865,
|
| 2079 |
+
"step": 26600
|
| 2080 |
+
},
|
| 2081 |
+
{
|
| 2082 |
+
"epoch": 0.8436125575046762,
|
| 2083 |
+
"grad_norm": 4.536484718322754,
|
| 2084 |
+
"learning_rate": 7.83753041170337e-06,
|
| 2085 |
+
"loss": 4.2786,
|
| 2086 |
+
"step": 26700
|
| 2087 |
+
},
|
| 2088 |
+
{
|
| 2089 |
+
"epoch": 0.8467721550983267,
|
| 2090 |
+
"grad_norm": 4.506146430969238,
|
| 2091 |
+
"learning_rate": 7.679547537046985e-06,
|
| 2092 |
+
"loss": 4.276,
|
| 2093 |
+
"step": 26800
|
| 2094 |
+
},
|
| 2095 |
+
{
|
| 2096 |
+
"epoch": 0.8499317526919772,
|
| 2097 |
+
"grad_norm": 5.557046413421631,
|
| 2098 |
+
"learning_rate": 7.521564662390597e-06,
|
| 2099 |
+
"loss": 4.2689,
|
| 2100 |
+
"step": 26900
|
| 2101 |
+
},
|
| 2102 |
+
{
|
| 2103 |
+
"epoch": 0.8530913502856277,
|
| 2104 |
+
"grad_norm": 4.260463237762451,
|
| 2105 |
+
"learning_rate": 7.365161616480774e-06,
|
| 2106 |
+
"loss": 4.2908,
|
| 2107 |
+
"step": 27000
|
| 2108 |
+
},
|
| 2109 |
+
{
|
| 2110 |
+
"epoch": 0.8530913502856277,
|
| 2111 |
+
"eval_loss": 4.402127265930176,
|
| 2112 |
+
"eval_runtime": 706.1851,
|
| 2113 |
+
"eval_samples_per_second": 124.494,
|
| 2114 |
+
"eval_steps_per_second": 15.562,
|
| 2115 |
+
"step": 27000
|
| 2116 |
+
},
|
| 2117 |
+
{
|
| 2118 |
+
"epoch": 0.8562509478792781,
|
| 2119 |
+
"grad_norm": 3.9850101470947266,
|
| 2120 |
+
"learning_rate": 7.207178741824388e-06,
|
| 2121 |
+
"loss": 4.2581,
|
| 2122 |
+
"step": 27100
|
| 2123 |
+
},
|
| 2124 |
+
{
|
| 2125 |
+
"epoch": 0.8594105454729286,
|
| 2126 |
+
"grad_norm": 5.228881359100342,
|
| 2127 |
+
"learning_rate": 7.049195867168e-06,
|
| 2128 |
+
"loss": 4.2347,
|
| 2129 |
+
"step": 27200
|
| 2130 |
+
},
|
| 2131 |
+
{
|
| 2132 |
+
"epoch": 0.8625701430665791,
|
| 2133 |
+
"grad_norm": 4.810471534729004,
|
| 2134 |
+
"learning_rate": 6.891212992511611e-06,
|
| 2135 |
+
"loss": 4.2821,
|
| 2136 |
+
"step": 27300
|
| 2137 |
+
},
|
| 2138 |
+
{
|
| 2139 |
+
"epoch": 0.8657297406602296,
|
| 2140 |
+
"grad_norm": 5.416518688201904,
|
| 2141 |
+
"learning_rate": 6.733230117855225e-06,
|
| 2142 |
+
"loss": 4.3175,
|
| 2143 |
+
"step": 27400
|
| 2144 |
+
},
|
| 2145 |
+
{
|
| 2146 |
+
"epoch": 0.86888933825388,
|
| 2147 |
+
"grad_norm": 4.266486167907715,
|
| 2148 |
+
"learning_rate": 6.575247243198837e-06,
|
| 2149 |
+
"loss": 4.219,
|
| 2150 |
+
"step": 27500
|
| 2151 |
+
},
|
| 2152 |
+
{
|
| 2153 |
+
"epoch": 0.8720489358475304,
|
| 2154 |
+
"grad_norm": 3.9196479320526123,
|
| 2155 |
+
"learning_rate": 6.41726436854245e-06,
|
| 2156 |
+
"loss": 4.2921,
|
| 2157 |
+
"step": 27600
|
| 2158 |
+
},
|
| 2159 |
+
{
|
| 2160 |
+
"epoch": 0.8752085334411809,
|
| 2161 |
+
"grad_norm": 3.8927292823791504,
|
| 2162 |
+
"learning_rate": 6.2592814938860626e-06,
|
| 2163 |
+
"loss": 4.224,
|
| 2164 |
+
"step": 27700
|
| 2165 |
+
},
|
| 2166 |
+
{
|
| 2167 |
+
"epoch": 0.8783681310348314,
|
| 2168 |
+
"grad_norm": 4.370504379272461,
|
| 2169 |
+
"learning_rate": 6.101298619229675e-06,
|
| 2170 |
+
"loss": 4.2426,
|
| 2171 |
+
"step": 27800
|
| 2172 |
+
},
|
| 2173 |
+
{
|
| 2174 |
+
"epoch": 0.8815277286284818,
|
| 2175 |
+
"grad_norm": 4.366906642913818,
|
| 2176 |
+
"learning_rate": 5.943315744573289e-06,
|
| 2177 |
+
"loss": 4.2668,
|
| 2178 |
+
"step": 27900
|
| 2179 |
+
},
|
| 2180 |
+
{
|
| 2181 |
+
"epoch": 0.8846873262221323,
|
| 2182 |
+
"grad_norm": 3.702138662338257,
|
| 2183 |
+
"learning_rate": 5.785332869916901e-06,
|
| 2184 |
+
"loss": 4.2331,
|
| 2185 |
+
"step": 28000
|
| 2186 |
+
},
|
| 2187 |
+
{
|
| 2188 |
+
"epoch": 0.8846873262221323,
|
| 2189 |
+
"eval_loss": 4.399970054626465,
|
| 2190 |
+
"eval_runtime": 705.6917,
|
| 2191 |
+
"eval_samples_per_second": 124.581,
|
| 2192 |
+
"eval_steps_per_second": 15.573,
|
| 2193 |
+
"step": 28000
|
| 2194 |
+
},
|
| 2195 |
+
{
|
| 2196 |
+
"epoch": 0.8878469238157828,
|
| 2197 |
+
"grad_norm": 3.696819543838501,
|
| 2198 |
+
"learning_rate": 5.627349995260514e-06,
|
| 2199 |
+
"loss": 4.2842,
|
| 2200 |
+
"step": 28100
|
| 2201 |
+
},
|
| 2202 |
+
{
|
| 2203 |
+
"epoch": 0.8910065214094333,
|
| 2204 |
+
"grad_norm": 3.837357759475708,
|
| 2205 |
+
"learning_rate": 5.469367120604127e-06,
|
| 2206 |
+
"loss": 4.2652,
|
| 2207 |
+
"step": 28200
|
| 2208 |
+
},
|
| 2209 |
+
{
|
| 2210 |
+
"epoch": 0.8941661190030837,
|
| 2211 |
+
"grad_norm": 4.607182502746582,
|
| 2212 |
+
"learning_rate": 5.31138424594774e-06,
|
| 2213 |
+
"loss": 4.2412,
|
| 2214 |
+
"step": 28300
|
| 2215 |
+
},
|
| 2216 |
+
{
|
| 2217 |
+
"epoch": 0.8973257165967342,
|
| 2218 |
+
"grad_norm": 3.803236484527588,
|
| 2219 |
+
"learning_rate": 5.154981200037916e-06,
|
| 2220 |
+
"loss": 4.2347,
|
| 2221 |
+
"step": 28400
|
| 2222 |
+
},
|
| 2223 |
+
{
|
| 2224 |
+
"epoch": 0.9004853141903847,
|
| 2225 |
+
"grad_norm": 4.182679653167725,
|
| 2226 |
+
"learning_rate": 4.9969983253815286e-06,
|
| 2227 |
+
"loss": 4.2815,
|
| 2228 |
+
"step": 28500
|
| 2229 |
+
},
|
| 2230 |
+
{
|
| 2231 |
+
"epoch": 0.9036449117840352,
|
| 2232 |
+
"grad_norm": 4.844381809234619,
|
| 2233 |
+
"learning_rate": 4.839015450725141e-06,
|
| 2234 |
+
"loss": 4.2565,
|
| 2235 |
+
"step": 28600
|
| 2236 |
+
},
|
| 2237 |
+
{
|
| 2238 |
+
"epoch": 0.9068045093776856,
|
| 2239 |
+
"grad_norm": 4.660447597503662,
|
| 2240 |
+
"learning_rate": 4.681032576068755e-06,
|
| 2241 |
+
"loss": 4.2364,
|
| 2242 |
+
"step": 28700
|
| 2243 |
+
},
|
| 2244 |
+
{
|
| 2245 |
+
"epoch": 0.9099641069713361,
|
| 2246 |
+
"grad_norm": 4.16555643081665,
|
| 2247 |
+
"learning_rate": 4.523049701412367e-06,
|
| 2248 |
+
"loss": 4.2775,
|
| 2249 |
+
"step": 28800
|
| 2250 |
+
},
|
| 2251 |
+
{
|
| 2252 |
+
"epoch": 0.9131237045649866,
|
| 2253 |
+
"grad_norm": 4.574762344360352,
|
| 2254 |
+
"learning_rate": 4.36506682675598e-06,
|
| 2255 |
+
"loss": 4.2255,
|
| 2256 |
+
"step": 28900
|
| 2257 |
+
},
|
| 2258 |
+
{
|
| 2259 |
+
"epoch": 0.9162833021586371,
|
| 2260 |
+
"grad_norm": 5.035285949707031,
|
| 2261 |
+
"learning_rate": 4.207083952099593e-06,
|
| 2262 |
+
"loss": 4.2787,
|
| 2263 |
+
"step": 29000
|
| 2264 |
+
},
|
| 2265 |
+
{
|
| 2266 |
+
"epoch": 0.9162833021586371,
|
| 2267 |
+
"eval_loss": 4.398350238800049,
|
| 2268 |
+
"eval_runtime": 702.9123,
|
| 2269 |
+
"eval_samples_per_second": 125.074,
|
| 2270 |
+
"eval_steps_per_second": 15.635,
|
| 2271 |
+
"step": 29000
|
| 2272 |
+
},
|
| 2273 |
+
{
|
| 2274 |
+
"epoch": 0.9194428997522875,
|
| 2275 |
+
"grad_norm": 4.223283290863037,
|
| 2276 |
+
"learning_rate": 4.049101077443205e-06,
|
| 2277 |
+
"loss": 4.2364,
|
| 2278 |
+
"step": 29100
|
| 2279 |
+
},
|
| 2280 |
+
{
|
| 2281 |
+
"epoch": 0.922602497345938,
|
| 2282 |
+
"grad_norm": 4.202021598815918,
|
| 2283 |
+
"learning_rate": 3.891118202786819e-06,
|
| 2284 |
+
"loss": 4.2258,
|
| 2285 |
+
"step": 29200
|
| 2286 |
+
},
|
| 2287 |
+
{
|
| 2288 |
+
"epoch": 0.9257620949395885,
|
| 2289 |
+
"grad_norm": 4.703751087188721,
|
| 2290 |
+
"learning_rate": 3.7331353281304305e-06,
|
| 2291 |
+
"loss": 4.2532,
|
| 2292 |
+
"step": 29300
|
| 2293 |
+
},
|
| 2294 |
+
{
|
| 2295 |
+
"epoch": 0.928921692533239,
|
| 2296 |
+
"grad_norm": 4.034895420074463,
|
| 2297 |
+
"learning_rate": 3.5751524534740436e-06,
|
| 2298 |
+
"loss": 4.2942,
|
| 2299 |
+
"step": 29400
|
| 2300 |
+
},
|
| 2301 |
+
{
|
| 2302 |
+
"epoch": 0.9320812901268895,
|
| 2303 |
+
"grad_norm": 5.5384521484375,
|
| 2304 |
+
"learning_rate": 3.417169578817656e-06,
|
| 2305 |
+
"loss": 4.1981,
|
| 2306 |
+
"step": 29500
|
| 2307 |
+
},
|
| 2308 |
+
{
|
| 2309 |
+
"epoch": 0.9352408877205399,
|
| 2310 |
+
"grad_norm": 4.258658409118652,
|
| 2311 |
+
"learning_rate": 3.2591867041612693e-06,
|
| 2312 |
+
"loss": 4.2552,
|
| 2313 |
+
"step": 29600
|
| 2314 |
+
},
|
| 2315 |
+
{
|
| 2316 |
+
"epoch": 0.9384004853141904,
|
| 2317 |
+
"grad_norm": 4.754565715789795,
|
| 2318 |
+
"learning_rate": 3.101203829504882e-06,
|
| 2319 |
+
"loss": 4.2769,
|
| 2320 |
+
"step": 29700
|
| 2321 |
+
},
|
| 2322 |
+
{
|
| 2323 |
+
"epoch": 0.9415600829078409,
|
| 2324 |
+
"grad_norm": 4.568323612213135,
|
| 2325 |
+
"learning_rate": 2.9432209548484945e-06,
|
| 2326 |
+
"loss": 4.2332,
|
| 2327 |
+
"step": 29800
|
| 2328 |
+
},
|
| 2329 |
+
{
|
| 2330 |
+
"epoch": 0.9447196805014914,
|
| 2331 |
+
"grad_norm": 4.62890625,
|
| 2332 |
+
"learning_rate": 2.785238080192107e-06,
|
| 2333 |
+
"loss": 4.2182,
|
| 2334 |
+
"step": 29900
|
| 2335 |
+
},
|
| 2336 |
+
{
|
| 2337 |
+
"epoch": 0.9478792780951418,
|
| 2338 |
+
"grad_norm": 4.388423919677734,
|
| 2339 |
+
"learning_rate": 2.6272552055357202e-06,
|
| 2340 |
+
"loss": 4.2405,
|
| 2341 |
+
"step": 30000
|
| 2342 |
+
},
|
| 2343 |
+
{
|
| 2344 |
+
"epoch": 0.9478792780951418,
|
| 2345 |
+
"eval_loss": 4.397016525268555,
|
| 2346 |
+
"eval_runtime": 702.9258,
|
| 2347 |
+
"eval_samples_per_second": 125.072,
|
| 2348 |
+
"eval_steps_per_second": 15.635,
|
| 2349 |
+
"step": 30000
|
| 2350 |
+
},
|
| 2351 |
+
{
|
| 2352 |
+
"epoch": 0.9510388756887923,
|
| 2353 |
+
"grad_norm": 4.26193904876709,
|
| 2354 |
+
"learning_rate": 2.469272330879333e-06,
|
| 2355 |
+
"loss": 4.2311,
|
| 2356 |
+
"step": 30100
|
| 2357 |
+
},
|
| 2358 |
+
{
|
| 2359 |
+
"epoch": 0.9541984732824428,
|
| 2360 |
+
"grad_norm": 4.310063362121582,
|
| 2361 |
+
"learning_rate": 2.3112894562229455e-06,
|
| 2362 |
+
"loss": 4.2259,
|
| 2363 |
+
"step": 30200
|
| 2364 |
+
},
|
| 2365 |
+
{
|
| 2366 |
+
"epoch": 0.9573580708760933,
|
| 2367 |
+
"grad_norm": 3.8391923904418945,
|
| 2368 |
+
"learning_rate": 2.153306581566558e-06,
|
| 2369 |
+
"loss": 4.224,
|
| 2370 |
+
"step": 30300
|
| 2371 |
+
},
|
| 2372 |
+
{
|
| 2373 |
+
"epoch": 0.9605176684697437,
|
| 2374 |
+
"grad_norm": 3.9689831733703613,
|
| 2375 |
+
"learning_rate": 1.995323706910171e-06,
|
| 2376 |
+
"loss": 4.2954,
|
| 2377 |
+
"step": 30400
|
| 2378 |
+
},
|
| 2379 |
+
{
|
| 2380 |
+
"epoch": 0.9636772660633942,
|
| 2381 |
+
"grad_norm": 3.916576623916626,
|
| 2382 |
+
"learning_rate": 1.8373408322537838e-06,
|
| 2383 |
+
"loss": 4.2109,
|
| 2384 |
+
"step": 30500
|
| 2385 |
+
},
|
| 2386 |
+
{
|
| 2387 |
+
"epoch": 0.9668368636570447,
|
| 2388 |
+
"grad_norm": 5.025672435760498,
|
| 2389 |
+
"learning_rate": 1.6793579575973967e-06,
|
| 2390 |
+
"loss": 4.2787,
|
| 2391 |
+
"step": 30600
|
| 2392 |
+
},
|
| 2393 |
+
{
|
| 2394 |
+
"epoch": 0.9699964612506952,
|
| 2395 |
+
"grad_norm": 4.102468013763428,
|
| 2396 |
+
"learning_rate": 1.5213750829410093e-06,
|
| 2397 |
+
"loss": 4.2559,
|
| 2398 |
+
"step": 30700
|
| 2399 |
+
},
|
| 2400 |
+
{
|
| 2401 |
+
"epoch": 0.9731560588443456,
|
| 2402 |
+
"grad_norm": 4.351522445678711,
|
| 2403 |
+
"learning_rate": 1.363392208284622e-06,
|
| 2404 |
+
"loss": 4.241,
|
| 2405 |
+
"step": 30800
|
| 2406 |
+
},
|
| 2407 |
+
{
|
| 2408 |
+
"epoch": 0.976315656437996,
|
| 2409 |
+
"grad_norm": 3.732164144515991,
|
| 2410 |
+
"learning_rate": 1.2054093336282348e-06,
|
| 2411 |
+
"loss": 4.2953,
|
| 2412 |
+
"step": 30900
|
| 2413 |
+
},
|
| 2414 |
+
{
|
| 2415 |
+
"epoch": 0.9794752540316465,
|
| 2416 |
+
"grad_norm": 4.7691969871521,
|
| 2417 |
+
"learning_rate": 1.0474264589718475e-06,
|
| 2418 |
+
"loss": 4.2715,
|
| 2419 |
+
"step": 31000
|
| 2420 |
+
},
|
| 2421 |
+
{
|
| 2422 |
+
"epoch": 0.9794752540316465,
|
| 2423 |
+
"eval_loss": 4.3962178230285645,
|
| 2424 |
+
"eval_runtime": 704.8954,
|
| 2425 |
+
"eval_samples_per_second": 124.722,
|
| 2426 |
+
"eval_steps_per_second": 15.591,
|
| 2427 |
+
"step": 31000
|
| 2428 |
+
},
|
| 2429 |
+
{
|
| 2430 |
+
"epoch": 0.982634851625297,
|
| 2431 |
+
"grad_norm": 4.083628177642822,
|
| 2432 |
+
"learning_rate": 8.894435843154603e-07,
|
| 2433 |
+
"loss": 4.2576,
|
| 2434 |
+
"step": 31100
|
| 2435 |
+
},
|
| 2436 |
+
{
|
| 2437 |
+
"epoch": 0.9857944492189474,
|
| 2438 |
+
"grad_norm": 4.615057945251465,
|
| 2439 |
+
"learning_rate": 7.314607096590729e-07,
|
| 2440 |
+
"loss": 4.271,
|
| 2441 |
+
"step": 31200
|
| 2442 |
+
},
|
| 2443 |
+
{
|
| 2444 |
+
"epoch": 0.9889540468125979,
|
| 2445 |
+
"grad_norm": 4.846735000610352,
|
| 2446 |
+
"learning_rate": 5.734778350026857e-07,
|
| 2447 |
+
"loss": 4.296,
|
| 2448 |
+
"step": 31300
|
| 2449 |
+
},
|
| 2450 |
+
{
|
| 2451 |
+
"epoch": 0.9921136444062484,
|
| 2452 |
+
"grad_norm": 4.390756130218506,
|
| 2453 |
+
"learning_rate": 4.154949603462984e-07,
|
| 2454 |
+
"loss": 4.247,
|
| 2455 |
+
"step": 31400
|
| 2456 |
+
},
|
| 2457 |
+
{
|
| 2458 |
+
"epoch": 0.9952732419998989,
|
| 2459 |
+
"grad_norm": 4.478681564331055,
|
| 2460 |
+
"learning_rate": 2.575120856899112e-07,
|
| 2461 |
+
"loss": 4.2517,
|
| 2462 |
+
"step": 31500
|
| 2463 |
+
},
|
| 2464 |
+
{
|
| 2465 |
+
"epoch": 0.9984328395935493,
|
| 2466 |
+
"grad_norm": 4.709827899932861,
|
| 2467 |
+
"learning_rate": 9.952921103352398e-08,
|
| 2468 |
+
"loss": 4.2248,
|
| 2469 |
+
"step": 31600
|
| 2470 |
+
}
|
| 2471 |
+
],
|
| 2472 |
+
"logging_steps": 100,
|
| 2473 |
+
"max_steps": 31649,
|
| 2474 |
+
"num_input_tokens_seen": 0,
|
| 2475 |
+
"num_train_epochs": 1,
|
| 2476 |
+
"save_steps": 10000,
|
| 2477 |
+
"stateful_callbacks": {
|
| 2478 |
+
"TrainerControl": {
|
| 2479 |
+
"args": {
|
| 2480 |
+
"should_epoch_stop": false,
|
| 2481 |
+
"should_evaluate": false,
|
| 2482 |
+
"should_log": false,
|
| 2483 |
+
"should_save": true,
|
| 2484 |
+
"should_training_stop": true
|
| 2485 |
+
},
|
| 2486 |
+
"attributes": {}
|
| 2487 |
+
}
|
| 2488 |
+
},
|
| 2489 |
+
"total_flos": 1.449889169178624e+17,
|
| 2490 |
+
"train_batch_size": 5,
|
| 2491 |
+
"trial_name": null,
|
| 2492 |
+
"trial_params": null
|
| 2493 |
+
}
|
fine_tuned_model_output_1/checkpoint-31649/training_args.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:95acebf5e801652183ac4ad483ff8bfe2ede739795ede7240bbee3b5a7707481
|
| 3 |
+
size 5368
|
fine_tuned_model_output_1/config.json
ADDED
|
@@ -0,0 +1,30 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"_name_or_path": "./linked_model_with_custom_tokenizer",
|
| 3 |
+
"architectures": [
|
| 4 |
+
"Qwen2ForCausalLM"
|
| 5 |
+
],
|
| 6 |
+
"attention_dropout": 0.0,
|
| 7 |
+
"bos_token_id": 151643,
|
| 8 |
+
"eos_token_id": 151643,
|
| 9 |
+
"hidden_act": "silu",
|
| 10 |
+
"hidden_size": 896,
|
| 11 |
+
"initializer_range": 0.02,
|
| 12 |
+
"intermediate_size": 4864,
|
| 13 |
+
"max_position_embeddings": 32768,
|
| 14 |
+
"max_window_layers": 24,
|
| 15 |
+
"model_type": "qwen2",
|
| 16 |
+
"num_attention_heads": 14,
|
| 17 |
+
"num_hidden_layers": 4,
|
| 18 |
+
"num_key_value_heads": 2,
|
| 19 |
+
"rms_norm_eps": 1e-06,
|
| 20 |
+
"rope_scaling": null,
|
| 21 |
+
"rope_theta": 1000000.0,
|
| 22 |
+
"sliding_window": 32768,
|
| 23 |
+
"tie_word_embeddings": true,
|
| 24 |
+
"torch_dtype": "float32",
|
| 25 |
+
"transformers_version": "4.49.0",
|
| 26 |
+
"use_cache": true,
|
| 27 |
+
"use_mrope": false,
|
| 28 |
+
"use_sliding_window": false,
|
| 29 |
+
"vocab_size": 32002
|
| 30 |
+
}
|
fine_tuned_model_output_1/generation_config.json
ADDED
|
@@ -0,0 +1,6 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"_from_model_config": true,
|
| 3 |
+
"bos_token_id": 151643,
|
| 4 |
+
"eos_token_id": 151643,
|
| 5 |
+
"transformers_version": "4.49.0"
|
| 6 |
+
}
|