Upload folder using huggingface_hub
Browse files- config.json +41 -0
- generation_config.json +6 -0
- optimizer.pt +3 -0
- pytorch_model.bin +3 -0
- rng_state.pth +3 -0
- scheduler.pt +3 -0
- trainer_state.json +3573 -0
- training_args.bin +3 -0
config.json
ADDED
|
@@ -0,0 +1,41 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"_name_or_path": "gpt2-medium",
|
| 3 |
+
"activation_function": "gelu_new",
|
| 4 |
+
"architectures": [
|
| 5 |
+
"GPT2LMHeadModel"
|
| 6 |
+
],
|
| 7 |
+
"attn_pdrop": 0.1,
|
| 8 |
+
"bos_token_id": 50256,
|
| 9 |
+
"embd_pdrop": 0.1,
|
| 10 |
+
"eos_token_id": 50256,
|
| 11 |
+
"initializer_range": 0.02,
|
| 12 |
+
"layer_norm_epsilon": 1e-05,
|
| 13 |
+
"model_type": "gpt2",
|
| 14 |
+
"n_ctx": 1024,
|
| 15 |
+
"n_embd": 1024,
|
| 16 |
+
"n_head": 16,
|
| 17 |
+
"n_inner": null,
|
| 18 |
+
"n_layer": 24,
|
| 19 |
+
"n_positions": 1024,
|
| 20 |
+
"n_special": 0,
|
| 21 |
+
"predict_special_tokens": true,
|
| 22 |
+
"reorder_and_upcast_attn": false,
|
| 23 |
+
"resid_pdrop": 0.1,
|
| 24 |
+
"scale_attn_by_inverse_layer_idx": false,
|
| 25 |
+
"scale_attn_weights": true,
|
| 26 |
+
"summary_activation": null,
|
| 27 |
+
"summary_first_dropout": 0.1,
|
| 28 |
+
"summary_proj_to_labels": true,
|
| 29 |
+
"summary_type": "cls_index",
|
| 30 |
+
"summary_use_proj": true,
|
| 31 |
+
"task_specific_params": {
|
| 32 |
+
"text-generation": {
|
| 33 |
+
"do_sample": true,
|
| 34 |
+
"max_length": 50
|
| 35 |
+
}
|
| 36 |
+
},
|
| 37 |
+
"torch_dtype": "float32",
|
| 38 |
+
"transformers_version": "4.48.1",
|
| 39 |
+
"use_cache": true,
|
| 40 |
+
"vocab_size": 50257
|
| 41 |
+
}
|
generation_config.json
ADDED
|
@@ -0,0 +1,6 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"_from_model_config": true,
|
| 3 |
+
"bos_token_id": 50256,
|
| 4 |
+
"eos_token_id": 50256,
|
| 5 |
+
"transformers_version": "4.48.1"
|
| 6 |
+
}
|
optimizer.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:db7739905c147b17d274771a25e55851f13f02174b812014d68a1a9cd62907e6
|
| 3 |
+
size 2838829242
|
pytorch_model.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:096fd587e795a89eeb51c8a0a424af48e0a357ece497f4554e415b0cdfd8ae59
|
| 3 |
+
size 1419388314
|
rng_state.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:059726101f5ec710d50a3b485f6858b22467df59fa4ff2ef558ac12ca72bea00
|
| 3 |
+
size 14244
|
scheduler.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:ac9d3e73e0c21c379b4e6670dc27c8dfe37b6b09e4c0969cc088d2ffcae7c045
|
| 3 |
+
size 1064
|
trainer_state.json
ADDED
|
@@ -0,0 +1,3573 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": null,
|
| 3 |
+
"best_model_checkpoint": null,
|
| 4 |
+
"epoch": 0.02869777005849354,
|
| 5 |
+
"eval_steps": 100,
|
| 6 |
+
"global_step": 500,
|
| 7 |
+
"is_hyper_param_search": false,
|
| 8 |
+
"is_local_process_zero": true,
|
| 9 |
+
"is_world_process_zero": true,
|
| 10 |
+
"log_history": [
|
| 11 |
+
{
|
| 12 |
+
"epoch": 5.739554011698708e-05,
|
| 13 |
+
"grad_norm": 1.8802112340927124,
|
| 14 |
+
"learning_rate": 5.9999999999999995e-05,
|
| 15 |
+
"loss": 2.9438,
|
| 16 |
+
"step": 1
|
| 17 |
+
},
|
| 18 |
+
{
|
| 19 |
+
"epoch": 0.00011479108023397416,
|
| 20 |
+
"grad_norm": 1.9408955574035645,
|
| 21 |
+
"learning_rate": 0.00011999999999999999,
|
| 22 |
+
"loss": 2.9429,
|
| 23 |
+
"step": 2
|
| 24 |
+
},
|
| 25 |
+
{
|
| 26 |
+
"epoch": 0.00017218662035096125,
|
| 27 |
+
"grad_norm": 2.9192652702331543,
|
| 28 |
+
"learning_rate": 0.00017999999999999998,
|
| 29 |
+
"loss": 2.952,
|
| 30 |
+
"step": 3
|
| 31 |
+
},
|
| 32 |
+
{
|
| 33 |
+
"epoch": 0.00022958216046794832,
|
| 34 |
+
"grad_norm": 2.3403642177581787,
|
| 35 |
+
"learning_rate": 0.00023999999999999998,
|
| 36 |
+
"loss": 2.9307,
|
| 37 |
+
"step": 4
|
| 38 |
+
},
|
| 39 |
+
{
|
| 40 |
+
"epoch": 0.00028697770058493544,
|
| 41 |
+
"grad_norm": 2.134683847427368,
|
| 42 |
+
"learning_rate": 0.0003,
|
| 43 |
+
"loss": 2.8917,
|
| 44 |
+
"step": 5
|
| 45 |
+
},
|
| 46 |
+
{
|
| 47 |
+
"epoch": 0.0003443732407019225,
|
| 48 |
+
"grad_norm": 1.5358260869979858,
|
| 49 |
+
"learning_rate": 0.00035999999999999997,
|
| 50 |
+
"loss": 2.9205,
|
| 51 |
+
"step": 6
|
| 52 |
+
},
|
| 53 |
+
{
|
| 54 |
+
"epoch": 0.0004017687808189096,
|
| 55 |
+
"grad_norm": 0.9012013673782349,
|
| 56 |
+
"learning_rate": 0.00041999999999999996,
|
| 57 |
+
"loss": 2.8937,
|
| 58 |
+
"step": 7
|
| 59 |
+
},
|
| 60 |
+
{
|
| 61 |
+
"epoch": 0.00045916432093589664,
|
| 62 |
+
"grad_norm": 0.9427694082260132,
|
| 63 |
+
"learning_rate": 0.00047999999999999996,
|
| 64 |
+
"loss": 2.904,
|
| 65 |
+
"step": 8
|
| 66 |
+
},
|
| 67 |
+
{
|
| 68 |
+
"epoch": 0.0005165598610528837,
|
| 69 |
+
"grad_norm": 1.662156105041504,
|
| 70 |
+
"learning_rate": 0.00054,
|
| 71 |
+
"loss": 2.9114,
|
| 72 |
+
"step": 9
|
| 73 |
+
},
|
| 74 |
+
{
|
| 75 |
+
"epoch": 0.0005739554011698709,
|
| 76 |
+
"grad_norm": 1.2877967357635498,
|
| 77 |
+
"learning_rate": 0.0006,
|
| 78 |
+
"loss": 2.9185,
|
| 79 |
+
"step": 10
|
| 80 |
+
},
|
| 81 |
+
{
|
| 82 |
+
"epoch": 0.000631350941286858,
|
| 83 |
+
"grad_norm": 1.3717082738876343,
|
| 84 |
+
"learning_rate": 0.0005999969170437548,
|
| 85 |
+
"loss": 2.899,
|
| 86 |
+
"step": 11
|
| 87 |
+
},
|
| 88 |
+
{
|
| 89 |
+
"epoch": 0.000688746481403845,
|
| 90 |
+
"grad_norm": 1.3706175088882446,
|
| 91 |
+
"learning_rate": 0.0005999876683017478,
|
| 92 |
+
"loss": 2.8522,
|
| 93 |
+
"step": 12
|
| 94 |
+
},
|
| 95 |
+
{
|
| 96 |
+
"epoch": 0.0007461420215208321,
|
| 97 |
+
"grad_norm": 0.7431464791297913,
|
| 98 |
+
"learning_rate": 0.0005999722541541584,
|
| 99 |
+
"loss": 2.8894,
|
| 100 |
+
"step": 13
|
| 101 |
+
},
|
| 102 |
+
{
|
| 103 |
+
"epoch": 0.0008035375616378192,
|
| 104 |
+
"grad_norm": 0.5839619040489197,
|
| 105 |
+
"learning_rate": 0.0005999506752346019,
|
| 106 |
+
"loss": 2.8866,
|
| 107 |
+
"step": 14
|
| 108 |
+
},
|
| 109 |
+
{
|
| 110 |
+
"epoch": 0.0008609331017548062,
|
| 111 |
+
"grad_norm": 0.5229901671409607,
|
| 112 |
+
"learning_rate": 0.0005999229324301031,
|
| 113 |
+
"loss": 2.8608,
|
| 114 |
+
"step": 15
|
| 115 |
+
},
|
| 116 |
+
{
|
| 117 |
+
"epoch": 0.0009183286418717933,
|
| 118 |
+
"grad_norm": 0.6879259943962097,
|
| 119 |
+
"learning_rate": 0.00059988902688106,
|
| 120 |
+
"loss": 2.8801,
|
| 121 |
+
"step": 16
|
| 122 |
+
},
|
| 123 |
+
{
|
| 124 |
+
"epoch": 0.0009757241819887805,
|
| 125 |
+
"grad_norm": 0.4949502646923065,
|
| 126 |
+
"learning_rate": 0.0005998489599811971,
|
| 127 |
+
"loss": 2.8857,
|
| 128 |
+
"step": 17
|
| 129 |
+
},
|
| 130 |
+
{
|
| 131 |
+
"epoch": 0.0010331197221057674,
|
| 132 |
+
"grad_norm": 0.5659216642379761,
|
| 133 |
+
"learning_rate": 0.0005998027333775077,
|
| 134 |
+
"loss": 2.8172,
|
| 135 |
+
"step": 18
|
| 136 |
+
},
|
| 137 |
+
{
|
| 138 |
+
"epoch": 0.0010905152622227546,
|
| 139 |
+
"grad_norm": 0.43849167227745056,
|
| 140 |
+
"learning_rate": 0.0005997503489701861,
|
| 141 |
+
"loss": 2.8479,
|
| 142 |
+
"step": 19
|
| 143 |
+
},
|
| 144 |
+
{
|
| 145 |
+
"epoch": 0.0011479108023397418,
|
| 146 |
+
"grad_norm": 0.5036750435829163,
|
| 147 |
+
"learning_rate": 0.0005996918089125504,
|
| 148 |
+
"loss": 2.8957,
|
| 149 |
+
"step": 20
|
| 150 |
+
},
|
| 151 |
+
{
|
| 152 |
+
"epoch": 0.0012053063424567287,
|
| 153 |
+
"grad_norm": 0.40093106031417847,
|
| 154 |
+
"learning_rate": 0.000599627115610953,
|
| 155 |
+
"loss": 2.8951,
|
| 156 |
+
"step": 21
|
| 157 |
+
},
|
| 158 |
+
{
|
| 159 |
+
"epoch": 0.001262701882573716,
|
| 160 |
+
"grad_norm": 0.3499244153499603,
|
| 161 |
+
"learning_rate": 0.0005995562717246821,
|
| 162 |
+
"loss": 2.8535,
|
| 163 |
+
"step": 22
|
| 164 |
+
},
|
| 165 |
+
{
|
| 166 |
+
"epoch": 0.0013200974226907029,
|
| 167 |
+
"grad_norm": 0.3672889769077301,
|
| 168 |
+
"learning_rate": 0.0005994792801658526,
|
| 169 |
+
"loss": 2.8507,
|
| 170 |
+
"step": 23
|
| 171 |
+
},
|
| 172 |
+
{
|
| 173 |
+
"epoch": 0.00137749296280769,
|
| 174 |
+
"grad_norm": 0.3307906985282898,
|
| 175 |
+
"learning_rate": 0.0005993961440992859,
|
| 176 |
+
"loss": 2.8597,
|
| 177 |
+
"step": 24
|
| 178 |
+
},
|
| 179 |
+
{
|
| 180 |
+
"epoch": 0.001434888502924677,
|
| 181 |
+
"grad_norm": 0.33352652192115784,
|
| 182 |
+
"learning_rate": 0.0005993068669423797,
|
| 183 |
+
"loss": 2.8023,
|
| 184 |
+
"step": 25
|
| 185 |
+
},
|
| 186 |
+
{
|
| 187 |
+
"epoch": 0.0014922840430416642,
|
| 188 |
+
"grad_norm": 0.30308255553245544,
|
| 189 |
+
"learning_rate": 0.0005992114523649685,
|
| 190 |
+
"loss": 2.864,
|
| 191 |
+
"step": 26
|
| 192 |
+
},
|
| 193 |
+
{
|
| 194 |
+
"epoch": 0.0015496795831586513,
|
| 195 |
+
"grad_norm": 0.2800331711769104,
|
| 196 |
+
"learning_rate": 0.000599109904289172,
|
| 197 |
+
"loss": 2.8459,
|
| 198 |
+
"step": 27
|
| 199 |
+
},
|
| 200 |
+
{
|
| 201 |
+
"epoch": 0.0016070751232756383,
|
| 202 |
+
"grad_norm": 0.2467849850654602,
|
| 203 |
+
"learning_rate": 0.0005990022268892337,
|
| 204 |
+
"loss": 2.8298,
|
| 205 |
+
"step": 28
|
| 206 |
+
},
|
| 207 |
+
{
|
| 208 |
+
"epoch": 0.0016644706633926255,
|
| 209 |
+
"grad_norm": 0.25928932428359985,
|
| 210 |
+
"learning_rate": 0.0005988884245913497,
|
| 211 |
+
"loss": 2.8061,
|
| 212 |
+
"step": 29
|
| 213 |
+
},
|
| 214 |
+
{
|
| 215 |
+
"epoch": 0.0017218662035096124,
|
| 216 |
+
"grad_norm": 0.2770285904407501,
|
| 217 |
+
"learning_rate": 0.0005987685020734869,
|
| 218 |
+
"loss": 2.8363,
|
| 219 |
+
"step": 30
|
| 220 |
+
},
|
| 221 |
+
{
|
| 222 |
+
"epoch": 0.0017792617436265996,
|
| 223 |
+
"grad_norm": 0.2888840436935425,
|
| 224 |
+
"learning_rate": 0.0005986424642651901,
|
| 225 |
+
"loss": 2.847,
|
| 226 |
+
"step": 31
|
| 227 |
+
},
|
| 228 |
+
{
|
| 229 |
+
"epoch": 0.0018366572837435866,
|
| 230 |
+
"grad_norm": 0.3389260172843933,
|
| 231 |
+
"learning_rate": 0.0005985103163473802,
|
| 232 |
+
"loss": 2.8185,
|
| 233 |
+
"step": 32
|
| 234 |
+
},
|
| 235 |
+
{
|
| 236 |
+
"epoch": 0.0018940528238605737,
|
| 237 |
+
"grad_norm": 0.3043622672557831,
|
| 238 |
+
"learning_rate": 0.0005983720637521404,
|
| 239 |
+
"loss": 2.8073,
|
| 240 |
+
"step": 33
|
| 241 |
+
},
|
| 242 |
+
{
|
| 243 |
+
"epoch": 0.001951448363977561,
|
| 244 |
+
"grad_norm": 0.2626359760761261,
|
| 245 |
+
"learning_rate": 0.0005982277121624933,
|
| 246 |
+
"loss": 2.8278,
|
| 247 |
+
"step": 34
|
| 248 |
+
},
|
| 249 |
+
{
|
| 250 |
+
"epoch": 0.002008843904094548,
|
| 251 |
+
"grad_norm": 0.2601317763328552,
|
| 252 |
+
"learning_rate": 0.0005980772675121675,
|
| 253 |
+
"loss": 2.8293,
|
| 254 |
+
"step": 35
|
| 255 |
+
},
|
| 256 |
+
{
|
| 257 |
+
"epoch": 0.002066239444211535,
|
| 258 |
+
"grad_norm": 0.2932066023349762,
|
| 259 |
+
"learning_rate": 0.0005979207359853532,
|
| 260 |
+
"loss": 2.842,
|
| 261 |
+
"step": 36
|
| 262 |
+
},
|
| 263 |
+
{
|
| 264 |
+
"epoch": 0.002123634984328522,
|
| 265 |
+
"grad_norm": 0.3828963041305542,
|
| 266 |
+
"learning_rate": 0.0005977581240164485,
|
| 267 |
+
"loss": 2.8383,
|
| 268 |
+
"step": 37
|
| 269 |
+
},
|
| 270 |
+
{
|
| 271 |
+
"epoch": 0.002181030524445509,
|
| 272 |
+
"grad_norm": 0.2928522527217865,
|
| 273 |
+
"learning_rate": 0.0005975894382897944,
|
| 274 |
+
"loss": 2.8291,
|
| 275 |
+
"step": 38
|
| 276 |
+
},
|
| 277 |
+
{
|
| 278 |
+
"epoch": 0.0022384260645624964,
|
| 279 |
+
"grad_norm": 0.2287234663963318,
|
| 280 |
+
"learning_rate": 0.0005974146857394005,
|
| 281 |
+
"loss": 2.8422,
|
| 282 |
+
"step": 39
|
| 283 |
+
},
|
| 284 |
+
{
|
| 285 |
+
"epoch": 0.0022958216046794835,
|
| 286 |
+
"grad_norm": 0.2722682058811188,
|
| 287 |
+
"learning_rate": 0.0005972338735486597,
|
| 288 |
+
"loss": 2.8217,
|
| 289 |
+
"step": 40
|
| 290 |
+
},
|
| 291 |
+
{
|
| 292 |
+
"epoch": 0.0023532171447964703,
|
| 293 |
+
"grad_norm": 0.21170516312122345,
|
| 294 |
+
"learning_rate": 0.0005970470091500531,
|
| 295 |
+
"loss": 2.831,
|
| 296 |
+
"step": 41
|
| 297 |
+
},
|
| 298 |
+
{
|
| 299 |
+
"epoch": 0.0024106126849134575,
|
| 300 |
+
"grad_norm": 0.22243160009384155,
|
| 301 |
+
"learning_rate": 0.0005968541002248439,
|
| 302 |
+
"loss": 2.862,
|
| 303 |
+
"step": 42
|
| 304 |
+
},
|
| 305 |
+
{
|
| 306 |
+
"epoch": 0.0024680082250304446,
|
| 307 |
+
"grad_norm": 0.18485133349895477,
|
| 308 |
+
"learning_rate": 0.0005966551547027627,
|
| 309 |
+
"loss": 2.8531,
|
| 310 |
+
"step": 43
|
| 311 |
+
},
|
| 312 |
+
{
|
| 313 |
+
"epoch": 0.002525403765147432,
|
| 314 |
+
"grad_norm": 0.21640127897262573,
|
| 315 |
+
"learning_rate": 0.0005964501807616806,
|
| 316 |
+
"loss": 2.8245,
|
| 317 |
+
"step": 44
|
| 318 |
+
},
|
| 319 |
+
{
|
| 320 |
+
"epoch": 0.0025827993052644185,
|
| 321 |
+
"grad_norm": 0.2716100513935089,
|
| 322 |
+
"learning_rate": 0.0005962391868272735,
|
| 323 |
+
"loss": 2.8093,
|
| 324 |
+
"step": 45
|
| 325 |
+
},
|
| 326 |
+
{
|
| 327 |
+
"epoch": 0.0026401948453814057,
|
| 328 |
+
"grad_norm": 0.19726517796516418,
|
| 329 |
+
"learning_rate": 0.0005960221815726757,
|
| 330 |
+
"loss": 2.8214,
|
| 331 |
+
"step": 46
|
| 332 |
+
},
|
| 333 |
+
{
|
| 334 |
+
"epoch": 0.002697590385498393,
|
| 335 |
+
"grad_norm": 0.2424098700284958,
|
| 336 |
+
"learning_rate": 0.0005957991739181231,
|
| 337 |
+
"loss": 2.818,
|
| 338 |
+
"step": 47
|
| 339 |
+
},
|
| 340 |
+
{
|
| 341 |
+
"epoch": 0.00275498592561538,
|
| 342 |
+
"grad_norm": 0.2414388209581375,
|
| 343 |
+
"learning_rate": 0.0005955701730305872,
|
| 344 |
+
"loss": 2.8491,
|
| 345 |
+
"step": 48
|
| 346 |
+
},
|
| 347 |
+
{
|
| 348 |
+
"epoch": 0.0028123814657323673,
|
| 349 |
+
"grad_norm": 0.25403571128845215,
|
| 350 |
+
"learning_rate": 0.0005953351883233972,
|
| 351 |
+
"loss": 2.8321,
|
| 352 |
+
"step": 49
|
| 353 |
+
},
|
| 354 |
+
{
|
| 355 |
+
"epoch": 0.002869777005849354,
|
| 356 |
+
"grad_norm": 0.30923786759376526,
|
| 357 |
+
"learning_rate": 0.0005950942294558544,
|
| 358 |
+
"loss": 2.8298,
|
| 359 |
+
"step": 50
|
| 360 |
+
},
|
| 361 |
+
{
|
| 362 |
+
"epoch": 0.002927172545966341,
|
| 363 |
+
"grad_norm": 0.22294141352176666,
|
| 364 |
+
"learning_rate": 0.0005948473063328338,
|
| 365 |
+
"loss": 2.8015,
|
| 366 |
+
"step": 51
|
| 367 |
+
},
|
| 368 |
+
{
|
| 369 |
+
"epoch": 0.0029845680860833283,
|
| 370 |
+
"grad_norm": 0.2882789075374603,
|
| 371 |
+
"learning_rate": 0.0005945944291043779,
|
| 372 |
+
"loss": 2.8256,
|
| 373 |
+
"step": 52
|
| 374 |
+
},
|
| 375 |
+
{
|
| 376 |
+
"epoch": 0.0030419636262003155,
|
| 377 |
+
"grad_norm": 0.25416064262390137,
|
| 378 |
+
"learning_rate": 0.0005943356081652793,
|
| 379 |
+
"loss": 2.8211,
|
| 380 |
+
"step": 53
|
| 381 |
+
},
|
| 382 |
+
{
|
| 383 |
+
"epoch": 0.0030993591663173027,
|
| 384 |
+
"grad_norm": 0.2488490343093872,
|
| 385 |
+
"learning_rate": 0.0005940708541546529,
|
| 386 |
+
"loss": 2.8618,
|
| 387 |
+
"step": 54
|
| 388 |
+
},
|
| 389 |
+
{
|
| 390 |
+
"epoch": 0.0031567547064342894,
|
| 391 |
+
"grad_norm": 0.27515849471092224,
|
| 392 |
+
"learning_rate": 0.000593800177955499,
|
| 393 |
+
"loss": 2.802,
|
| 394 |
+
"step": 55
|
| 395 |
+
},
|
| 396 |
+
{
|
| 397 |
+
"epoch": 0.0032141502465512766,
|
| 398 |
+
"grad_norm": 0.2030380666255951,
|
| 399 |
+
"learning_rate": 0.0005935235906942563,
|
| 400 |
+
"loss": 2.8229,
|
| 401 |
+
"step": 56
|
| 402 |
+
},
|
| 403 |
+
{
|
| 404 |
+
"epoch": 0.003271545786668264,
|
| 405 |
+
"grad_norm": 0.2384052276611328,
|
| 406 |
+
"learning_rate": 0.0005932411037403436,
|
| 407 |
+
"loss": 2.8122,
|
| 408 |
+
"step": 57
|
| 409 |
+
},
|
| 410 |
+
{
|
| 411 |
+
"epoch": 0.003328941326785251,
|
| 412 |
+
"grad_norm": 0.2543489336967468,
|
| 413 |
+
"learning_rate": 0.000592952728705693,
|
| 414 |
+
"loss": 2.8302,
|
| 415 |
+
"step": 58
|
| 416 |
+
},
|
| 417 |
+
{
|
| 418 |
+
"epoch": 0.003386336866902238,
|
| 419 |
+
"grad_norm": 0.2387794405221939,
|
| 420 |
+
"learning_rate": 0.000592658477444273,
|
| 421 |
+
"loss": 2.835,
|
| 422 |
+
"step": 59
|
| 423 |
+
},
|
| 424 |
+
{
|
| 425 |
+
"epoch": 0.003443732407019225,
|
| 426 |
+
"grad_norm": 0.2748169004917145,
|
| 427 |
+
"learning_rate": 0.0005923583620516003,
|
| 428 |
+
"loss": 2.834,
|
| 429 |
+
"step": 60
|
| 430 |
+
},
|
| 431 |
+
{
|
| 432 |
+
"epoch": 0.003501127947136212,
|
| 433 |
+
"grad_norm": 0.2565017640590668,
|
| 434 |
+
"learning_rate": 0.0005920523948642431,
|
| 435 |
+
"loss": 2.8452,
|
| 436 |
+
"step": 61
|
| 437 |
+
},
|
| 438 |
+
{
|
| 439 |
+
"epoch": 0.0035585234872531992,
|
| 440 |
+
"grad_norm": 0.25502678751945496,
|
| 441 |
+
"learning_rate": 0.0005917405884593144,
|
| 442 |
+
"loss": 2.8345,
|
| 443 |
+
"step": 62
|
| 444 |
+
},
|
| 445 |
+
{
|
| 446 |
+
"epoch": 0.0036159190273701864,
|
| 447 |
+
"grad_norm": 0.22830121219158173,
|
| 448 |
+
"learning_rate": 0.0005914229556539538,
|
| 449 |
+
"loss": 2.7989,
|
| 450 |
+
"step": 63
|
| 451 |
+
},
|
| 452 |
+
{
|
| 453 |
+
"epoch": 0.003673314567487173,
|
| 454 |
+
"grad_norm": 0.3146669268608093,
|
| 455 |
+
"learning_rate": 0.0005910995095048024,
|
| 456 |
+
"loss": 2.845,
|
| 457 |
+
"step": 64
|
| 458 |
+
},
|
| 459 |
+
{
|
| 460 |
+
"epoch": 0.0037307101076041603,
|
| 461 |
+
"grad_norm": 0.2924383580684662,
|
| 462 |
+
"learning_rate": 0.000590770263307464,
|
| 463 |
+
"loss": 2.8303,
|
| 464 |
+
"step": 65
|
| 465 |
+
},
|
| 466 |
+
{
|
| 467 |
+
"epoch": 0.0037881056477211475,
|
| 468 |
+
"grad_norm": 0.2577711343765259,
|
| 469 |
+
"learning_rate": 0.0005904352305959605,
|
| 470 |
+
"loss": 2.8156,
|
| 471 |
+
"step": 66
|
| 472 |
+
},
|
| 473 |
+
{
|
| 474 |
+
"epoch": 0.0038455011878381347,
|
| 475 |
+
"grad_norm": 0.2631978988647461,
|
| 476 |
+
"learning_rate": 0.0005900944251421745,
|
| 477 |
+
"loss": 2.833,
|
| 478 |
+
"step": 67
|
| 479 |
+
},
|
| 480 |
+
{
|
| 481 |
+
"epoch": 0.003902896727955122,
|
| 482 |
+
"grad_norm": 0.21994397044181824,
|
| 483 |
+
"learning_rate": 0.000589747860955283,
|
| 484 |
+
"loss": 2.8136,
|
| 485 |
+
"step": 68
|
| 486 |
+
},
|
| 487 |
+
{
|
| 488 |
+
"epoch": 0.003960292268072109,
|
| 489 |
+
"grad_norm": 0.3000943064689636,
|
| 490 |
+
"learning_rate": 0.0005893955522811827,
|
| 491 |
+
"loss": 2.8415,
|
| 492 |
+
"step": 69
|
| 493 |
+
},
|
| 494 |
+
{
|
| 495 |
+
"epoch": 0.004017687808189096,
|
| 496 |
+
"grad_norm": 0.24310976266860962,
|
| 497 |
+
"learning_rate": 0.0005890375136019032,
|
| 498 |
+
"loss": 2.8148,
|
| 499 |
+
"step": 70
|
| 500 |
+
},
|
| 501 |
+
{
|
| 502 |
+
"epoch": 0.004075083348306083,
|
| 503 |
+
"grad_norm": 0.24616850912570953,
|
| 504 |
+
"learning_rate": 0.0005886737596350122,
|
| 505 |
+
"loss": 2.8329,
|
| 506 |
+
"step": 71
|
| 507 |
+
},
|
| 508 |
+
{
|
| 509 |
+
"epoch": 0.00413247888842307,
|
| 510 |
+
"grad_norm": 0.2714521884918213,
|
| 511 |
+
"learning_rate": 0.0005883043053330105,
|
| 512 |
+
"loss": 2.8356,
|
| 513 |
+
"step": 72
|
| 514 |
+
},
|
| 515 |
+
{
|
| 516 |
+
"epoch": 0.004189874428540057,
|
| 517 |
+
"grad_norm": 0.2601388096809387,
|
| 518 |
+
"learning_rate": 0.0005879291658827176,
|
| 519 |
+
"loss": 2.8228,
|
| 520 |
+
"step": 73
|
| 521 |
+
},
|
| 522 |
+
{
|
| 523 |
+
"epoch": 0.004247269968657044,
|
| 524 |
+
"grad_norm": 0.22764116525650024,
|
| 525 |
+
"learning_rate": 0.0005875483567046467,
|
| 526 |
+
"loss": 2.801,
|
| 527 |
+
"step": 74
|
| 528 |
+
},
|
| 529 |
+
{
|
| 530 |
+
"epoch": 0.004304665508774032,
|
| 531 |
+
"grad_norm": 0.22346433997154236,
|
| 532 |
+
"learning_rate": 0.0005871618934523719,
|
| 533 |
+
"loss": 2.7948,
|
| 534 |
+
"step": 75
|
| 535 |
+
},
|
| 536 |
+
{
|
| 537 |
+
"epoch": 0.004362061048891018,
|
| 538 |
+
"grad_norm": 0.18839874863624573,
|
| 539 |
+
"learning_rate": 0.0005867697920118835,
|
| 540 |
+
"loss": 2.8341,
|
| 541 |
+
"step": 76
|
| 542 |
+
},
|
| 543 |
+
{
|
| 544 |
+
"epoch": 0.004419456589008005,
|
| 545 |
+
"grad_norm": 0.25794312357902527,
|
| 546 |
+
"learning_rate": 0.0005863720685009362,
|
| 547 |
+
"loss": 2.815,
|
| 548 |
+
"step": 77
|
| 549 |
+
},
|
| 550 |
+
{
|
| 551 |
+
"epoch": 0.004476852129124993,
|
| 552 |
+
"grad_norm": 0.2352106124162674,
|
| 553 |
+
"learning_rate": 0.0005859687392683856,
|
| 554 |
+
"loss": 2.8169,
|
| 555 |
+
"step": 78
|
| 556 |
+
},
|
| 557 |
+
{
|
| 558 |
+
"epoch": 0.0045342476692419795,
|
| 559 |
+
"grad_norm": 0.28784099221229553,
|
| 560 |
+
"learning_rate": 0.0005855598208935169,
|
| 561 |
+
"loss": 2.8506,
|
| 562 |
+
"step": 79
|
| 563 |
+
},
|
| 564 |
+
{
|
| 565 |
+
"epoch": 0.004591643209358967,
|
| 566 |
+
"grad_norm": 0.22999855875968933,
|
| 567 |
+
"learning_rate": 0.0005851453301853628,
|
| 568 |
+
"loss": 2.8377,
|
| 569 |
+
"step": 80
|
| 570 |
+
},
|
| 571 |
+
{
|
| 572 |
+
"epoch": 0.004649038749475954,
|
| 573 |
+
"grad_norm": 0.21411263942718506,
|
| 574 |
+
"learning_rate": 0.0005847252841820128,
|
| 575 |
+
"loss": 2.8137,
|
| 576 |
+
"step": 81
|
| 577 |
+
},
|
| 578 |
+
{
|
| 579 |
+
"epoch": 0.0047064342895929406,
|
| 580 |
+
"grad_norm": 0.2420736700296402,
|
| 581 |
+
"learning_rate": 0.0005842997001499129,
|
| 582 |
+
"loss": 2.7929,
|
| 583 |
+
"step": 82
|
| 584 |
+
},
|
| 585 |
+
{
|
| 586 |
+
"epoch": 0.004763829829709928,
|
| 587 |
+
"grad_norm": 0.24426190555095673,
|
| 588 |
+
"learning_rate": 0.0005838685955831558,
|
| 589 |
+
"loss": 2.8273,
|
| 590 |
+
"step": 83
|
| 591 |
+
},
|
| 592 |
+
{
|
| 593 |
+
"epoch": 0.004821225369826915,
|
| 594 |
+
"grad_norm": 0.20297811925411224,
|
| 595 |
+
"learning_rate": 0.0005834319882027617,
|
| 596 |
+
"loss": 2.7993,
|
| 597 |
+
"step": 84
|
| 598 |
+
},
|
| 599 |
+
{
|
| 600 |
+
"epoch": 0.0048786209099439025,
|
| 601 |
+
"grad_norm": 0.2474389523267746,
|
| 602 |
+
"learning_rate": 0.00058298989595595,
|
| 603 |
+
"loss": 2.8252,
|
| 604 |
+
"step": 85
|
| 605 |
+
},
|
| 606 |
+
{
|
| 607 |
+
"epoch": 0.004936016450060889,
|
| 608 |
+
"grad_norm": 0.22601982951164246,
|
| 609 |
+
"learning_rate": 0.0005825423370154012,
|
| 610 |
+
"loss": 2.8421,
|
| 611 |
+
"step": 86
|
| 612 |
+
},
|
| 613 |
+
{
|
| 614 |
+
"epoch": 0.004993411990177876,
|
| 615 |
+
"grad_norm": 0.24997788667678833,
|
| 616 |
+
"learning_rate": 0.0005820893297785106,
|
| 617 |
+
"loss": 2.8485,
|
| 618 |
+
"step": 87
|
| 619 |
+
},
|
| 620 |
+
{
|
| 621 |
+
"epoch": 0.005050807530294864,
|
| 622 |
+
"grad_norm": 0.19994623959064484,
|
| 623 |
+
"learning_rate": 0.0005816308928666314,
|
| 624 |
+
"loss": 2.8456,
|
| 625 |
+
"step": 88
|
| 626 |
+
},
|
| 627 |
+
{
|
| 628 |
+
"epoch": 0.00510820307041185,
|
| 629 |
+
"grad_norm": 0.19206245243549347,
|
| 630 |
+
"learning_rate": 0.0005811670451243093,
|
| 631 |
+
"loss": 2.8035,
|
| 632 |
+
"step": 89
|
| 633 |
+
},
|
| 634 |
+
{
|
| 635 |
+
"epoch": 0.005165598610528837,
|
| 636 |
+
"grad_norm": 0.2515026032924652,
|
| 637 |
+
"learning_rate": 0.0005806978056185083,
|
| 638 |
+
"loss": 2.8232,
|
| 639 |
+
"step": 90
|
| 640 |
+
},
|
| 641 |
+
{
|
| 642 |
+
"epoch": 0.005222994150645825,
|
| 643 |
+
"grad_norm": 0.22921022772789001,
|
| 644 |
+
"learning_rate": 0.0005802231936378267,
|
| 645 |
+
"loss": 2.8366,
|
| 646 |
+
"step": 91
|
| 647 |
+
},
|
| 648 |
+
{
|
| 649 |
+
"epoch": 0.0052803896907628114,
|
| 650 |
+
"grad_norm": 0.248809352517128,
|
| 651 |
+
"learning_rate": 0.000579743228691704,
|
| 652 |
+
"loss": 2.8331,
|
| 653 |
+
"step": 92
|
| 654 |
+
},
|
| 655 |
+
{
|
| 656 |
+
"epoch": 0.005337785230879799,
|
| 657 |
+
"grad_norm": 0.18247073888778687,
|
| 658 |
+
"learning_rate": 0.0005792579305096191,
|
| 659 |
+
"loss": 2.8249,
|
| 660 |
+
"step": 93
|
| 661 |
+
},
|
| 662 |
+
{
|
| 663 |
+
"epoch": 0.005395180770996786,
|
| 664 |
+
"grad_norm": 0.2440440058708191,
|
| 665 |
+
"learning_rate": 0.0005787673190402799,
|
| 666 |
+
"loss": 2.837,
|
| 667 |
+
"step": 94
|
| 668 |
+
},
|
| 669 |
+
{
|
| 670 |
+
"epoch": 0.0054525763111137725,
|
| 671 |
+
"grad_norm": 0.21160444617271423,
|
| 672 |
+
"learning_rate": 0.0005782714144508019,
|
| 673 |
+
"loss": 2.7864,
|
| 674 |
+
"step": 95
|
| 675 |
+
},
|
| 676 |
+
{
|
| 677 |
+
"epoch": 0.00550997185123076,
|
| 678 |
+
"grad_norm": 0.21344538033008575,
|
| 679 |
+
"learning_rate": 0.0005777702371258806,
|
| 680 |
+
"loss": 2.847,
|
| 681 |
+
"step": 96
|
| 682 |
+
},
|
| 683 |
+
{
|
| 684 |
+
"epoch": 0.005567367391347747,
|
| 685 |
+
"grad_norm": 0.24861139059066772,
|
| 686 |
+
"learning_rate": 0.0005772638076669529,
|
| 687 |
+
"loss": 2.8267,
|
| 688 |
+
"step": 97
|
| 689 |
+
},
|
| 690 |
+
{
|
| 691 |
+
"epoch": 0.0056247629314647345,
|
| 692 |
+
"grad_norm": 0.290520042181015,
|
| 693 |
+
"learning_rate": 0.0005767521468913501,
|
| 694 |
+
"loss": 2.827,
|
| 695 |
+
"step": 98
|
| 696 |
+
},
|
| 697 |
+
{
|
| 698 |
+
"epoch": 0.005682158471581721,
|
| 699 |
+
"grad_norm": 0.20536312460899353,
|
| 700 |
+
"learning_rate": 0.0005762352758314429,
|
| 701 |
+
"loss": 2.8476,
|
| 702 |
+
"step": 99
|
| 703 |
+
},
|
| 704 |
+
{
|
| 705 |
+
"epoch": 0.005739554011698708,
|
| 706 |
+
"grad_norm": 0.21782469749450684,
|
| 707 |
+
"learning_rate": 0.000575713215733776,
|
| 708 |
+
"loss": 2.844,
|
| 709 |
+
"step": 100
|
| 710 |
+
},
|
| 711 |
+
{
|
| 712 |
+
"epoch": 0.005739554011698708,
|
| 713 |
+
"eval_loss": 2.7509028911590576,
|
| 714 |
+
"eval_runtime": 85.2068,
|
| 715 |
+
"eval_samples_per_second": 50.641,
|
| 716 |
+
"eval_steps_per_second": 12.663,
|
| 717 |
+
"step": 100
|
| 718 |
+
},
|
| 719 |
+
{
|
| 720 |
+
"epoch": 0.005796949551815696,
|
| 721 |
+
"grad_norm": 0.2523731291294098,
|
| 722 |
+
"learning_rate": 0.0005751859880581954,
|
| 723 |
+
"loss": 2.8125,
|
| 724 |
+
"step": 101
|
| 725 |
+
},
|
| 726 |
+
{
|
| 727 |
+
"epoch": 0.005854345091932682,
|
| 728 |
+
"grad_norm": 0.30107325315475464,
|
| 729 |
+
"learning_rate": 0.0005746536144769656,
|
| 730 |
+
"loss": 2.8108,
|
| 731 |
+
"step": 102
|
| 732 |
+
},
|
| 733 |
+
{
|
| 734 |
+
"epoch": 0.00591174063204967,
|
| 735 |
+
"grad_norm": 0.24103832244873047,
|
| 736 |
+
"learning_rate": 0.0005741161168738794,
|
| 737 |
+
"loss": 2.8282,
|
| 738 |
+
"step": 103
|
| 739 |
+
},
|
| 740 |
+
{
|
| 741 |
+
"epoch": 0.005969136172166657,
|
| 742 |
+
"grad_norm": 0.31273001432418823,
|
| 743 |
+
"learning_rate": 0.0005735735173433582,
|
| 744 |
+
"loss": 2.8104,
|
| 745 |
+
"step": 104
|
| 746 |
+
},
|
| 747 |
+
{
|
| 748 |
+
"epoch": 0.006026531712283643,
|
| 749 |
+
"grad_norm": 0.19059035181999207,
|
| 750 |
+
"learning_rate": 0.0005730258381895433,
|
| 751 |
+
"loss": 2.8186,
|
| 752 |
+
"step": 105
|
| 753 |
+
},
|
| 754 |
+
{
|
| 755 |
+
"epoch": 0.006083927252400631,
|
| 756 |
+
"grad_norm": 0.25082021951675415,
|
| 757 |
+
"learning_rate": 0.0005724731019253797,
|
| 758 |
+
"loss": 2.8154,
|
| 759 |
+
"step": 106
|
| 760 |
+
},
|
| 761 |
+
{
|
| 762 |
+
"epoch": 0.006141322792517618,
|
| 763 |
+
"grad_norm": 0.23254480957984924,
|
| 764 |
+
"learning_rate": 0.0005719153312716904,
|
| 765 |
+
"loss": 2.8121,
|
| 766 |
+
"step": 107
|
| 767 |
+
},
|
| 768 |
+
{
|
| 769 |
+
"epoch": 0.006198718332634605,
|
| 770 |
+
"grad_norm": 0.24095705151557922,
|
| 771 |
+
"learning_rate": 0.0005713525491562421,
|
| 772 |
+
"loss": 2.8361,
|
| 773 |
+
"step": 108
|
| 774 |
+
},
|
| 775 |
+
{
|
| 776 |
+
"epoch": 0.006256113872751592,
|
| 777 |
+
"grad_norm": 0.17760275304317474,
|
| 778 |
+
"learning_rate": 0.0005707847787128034,
|
| 779 |
+
"loss": 2.8396,
|
| 780 |
+
"step": 109
|
| 781 |
+
},
|
| 782 |
+
{
|
| 783 |
+
"epoch": 0.006313509412868579,
|
| 784 |
+
"grad_norm": 0.20905229449272156,
|
| 785 |
+
"learning_rate": 0.0005702120432801934,
|
| 786 |
+
"loss": 2.8284,
|
| 787 |
+
"step": 110
|
| 788 |
+
},
|
| 789 |
+
{
|
| 790 |
+
"epoch": 0.0063709049529855665,
|
| 791 |
+
"grad_norm": 0.19538630545139313,
|
| 792 |
+
"learning_rate": 0.0005696343664013227,
|
| 793 |
+
"loss": 2.8417,
|
| 794 |
+
"step": 111
|
| 795 |
+
},
|
| 796 |
+
{
|
| 797 |
+
"epoch": 0.006428300493102553,
|
| 798 |
+
"grad_norm": 0.2408672571182251,
|
| 799 |
+
"learning_rate": 0.0005690517718222248,
|
| 800 |
+
"loss": 2.8416,
|
| 801 |
+
"step": 112
|
| 802 |
+
},
|
| 803 |
+
{
|
| 804 |
+
"epoch": 0.006485696033219541,
|
| 805 |
+
"grad_norm": 0.19618412852287292,
|
| 806 |
+
"learning_rate": 0.0005684642834910813,
|
| 807 |
+
"loss": 2.8683,
|
| 808 |
+
"step": 113
|
| 809 |
+
},
|
| 810 |
+
{
|
| 811 |
+
"epoch": 0.006543091573336528,
|
| 812 |
+
"grad_norm": 0.17854906618595123,
|
| 813 |
+
"learning_rate": 0.0005678719255572363,
|
| 814 |
+
"loss": 2.8232,
|
| 815 |
+
"step": 114
|
| 816 |
+
},
|
| 817 |
+
{
|
| 818 |
+
"epoch": 0.006600487113453514,
|
| 819 |
+
"grad_norm": 0.2527766227722168,
|
| 820 |
+
"learning_rate": 0.0005672747223702044,
|
| 821 |
+
"loss": 2.8219,
|
| 822 |
+
"step": 115
|
| 823 |
+
},
|
| 824 |
+
{
|
| 825 |
+
"epoch": 0.006657882653570502,
|
| 826 |
+
"grad_norm": 0.21465440094470978,
|
| 827 |
+
"learning_rate": 0.0005666726984786695,
|
| 828 |
+
"loss": 2.8308,
|
| 829 |
+
"step": 116
|
| 830 |
+
},
|
| 831 |
+
{
|
| 832 |
+
"epoch": 0.006715278193687489,
|
| 833 |
+
"grad_norm": 0.2080729454755783,
|
| 834 |
+
"learning_rate": 0.000566065878629476,
|
| 835 |
+
"loss": 2.8369,
|
| 836 |
+
"step": 117
|
| 837 |
+
},
|
| 838 |
+
{
|
| 839 |
+
"epoch": 0.006772673733804476,
|
| 840 |
+
"grad_norm": 0.18979360163211823,
|
| 841 |
+
"learning_rate": 0.0005654542877666108,
|
| 842 |
+
"loss": 2.7997,
|
| 843 |
+
"step": 118
|
| 844 |
+
},
|
| 845 |
+
{
|
| 846 |
+
"epoch": 0.006830069273921463,
|
| 847 |
+
"grad_norm": 0.20258580148220062,
|
| 848 |
+
"learning_rate": 0.0005648379510301792,
|
| 849 |
+
"loss": 2.846,
|
| 850 |
+
"step": 119
|
| 851 |
+
},
|
| 852 |
+
{
|
| 853 |
+
"epoch": 0.00688746481403845,
|
| 854 |
+
"grad_norm": 0.2112026810646057,
|
| 855 |
+
"learning_rate": 0.0005642168937553701,
|
| 856 |
+
"loss": 2.8521,
|
| 857 |
+
"step": 120
|
| 858 |
+
},
|
| 859 |
+
{
|
| 860 |
+
"epoch": 0.006944860354155437,
|
| 861 |
+
"grad_norm": 0.25105029344558716,
|
| 862 |
+
"learning_rate": 0.0005635911414714158,
|
| 863 |
+
"loss": 2.8081,
|
| 864 |
+
"step": 121
|
| 865 |
+
},
|
| 866 |
+
{
|
| 867 |
+
"epoch": 0.007002255894272424,
|
| 868 |
+
"grad_norm": 0.21830224990844727,
|
| 869 |
+
"learning_rate": 0.0005629607199005416,
|
| 870 |
+
"loss": 2.8161,
|
| 871 |
+
"step": 122
|
| 872 |
+
},
|
| 873 |
+
{
|
| 874 |
+
"epoch": 0.007059651434389411,
|
| 875 |
+
"grad_norm": 0.19216330349445343,
|
| 876 |
+
"learning_rate": 0.0005623256549569091,
|
| 877 |
+
"loss": 2.805,
|
| 878 |
+
"step": 123
|
| 879 |
+
},
|
| 880 |
+
{
|
| 881 |
+
"epoch": 0.0071170469745063985,
|
| 882 |
+
"grad_norm": 0.19969609379768372,
|
| 883 |
+
"learning_rate": 0.000561685972745551,
|
| 884 |
+
"loss": 2.7859,
|
| 885 |
+
"step": 124
|
| 886 |
+
},
|
| 887 |
+
{
|
| 888 |
+
"epoch": 0.007174442514623385,
|
| 889 |
+
"grad_norm": 0.22093947231769562,
|
| 890 |
+
"learning_rate": 0.0005610416995612973,
|
| 891 |
+
"loss": 2.8194,
|
| 892 |
+
"step": 125
|
| 893 |
+
},
|
| 894 |
+
{
|
| 895 |
+
"epoch": 0.007231838054740373,
|
| 896 |
+
"grad_norm": 0.2148187905550003,
|
| 897 |
+
"learning_rate": 0.0005603928618876952,
|
| 898 |
+
"loss": 2.8565,
|
| 899 |
+
"step": 126
|
| 900 |
+
},
|
| 901 |
+
{
|
| 902 |
+
"epoch": 0.0072892335948573595,
|
| 903 |
+
"grad_norm": 0.18277674913406372,
|
| 904 |
+
"learning_rate": 0.0005597394863959201,
|
| 905 |
+
"loss": 2.8187,
|
| 906 |
+
"step": 127
|
| 907 |
+
},
|
| 908 |
+
{
|
| 909 |
+
"epoch": 0.007346629134974346,
|
| 910 |
+
"grad_norm": 0.22607837617397308,
|
| 911 |
+
"learning_rate": 0.0005590815999436795,
|
| 912 |
+
"loss": 2.8607,
|
| 913 |
+
"step": 128
|
| 914 |
+
},
|
| 915 |
+
{
|
| 916 |
+
"epoch": 0.007404024675091334,
|
| 917 |
+
"grad_norm": 0.22417186200618744,
|
| 918 |
+
"learning_rate": 0.0005584192295741086,
|
| 919 |
+
"loss": 2.8198,
|
| 920 |
+
"step": 129
|
| 921 |
+
},
|
| 922 |
+
{
|
| 923 |
+
"epoch": 0.007461420215208321,
|
| 924 |
+
"grad_norm": 0.229670912027359,
|
| 925 |
+
"learning_rate": 0.0005577524025146591,
|
| 926 |
+
"loss": 2.8477,
|
| 927 |
+
"step": 130
|
| 928 |
+
},
|
| 929 |
+
{
|
| 930 |
+
"epoch": 0.007518815755325308,
|
| 931 |
+
"grad_norm": 0.1985808163881302,
|
| 932 |
+
"learning_rate": 0.0005570811461759794,
|
| 933 |
+
"loss": 2.8058,
|
| 934 |
+
"step": 131
|
| 935 |
+
},
|
| 936 |
+
{
|
| 937 |
+
"epoch": 0.007576211295442295,
|
| 938 |
+
"grad_norm": 0.22260330617427826,
|
| 939 |
+
"learning_rate": 0.0005564054881507886,
|
| 940 |
+
"loss": 2.8369,
|
| 941 |
+
"step": 132
|
| 942 |
+
},
|
| 943 |
+
{
|
| 944 |
+
"epoch": 0.007633606835559282,
|
| 945 |
+
"grad_norm": 0.20925524830818176,
|
| 946 |
+
"learning_rate": 0.0005557254562127417,
|
| 947 |
+
"loss": 2.8205,
|
| 948 |
+
"step": 133
|
| 949 |
+
},
|
| 950 |
+
{
|
| 951 |
+
"epoch": 0.007691002375676269,
|
| 952 |
+
"grad_norm": 0.26581674814224243,
|
| 953 |
+
"learning_rate": 0.0005550410783152882,
|
| 954 |
+
"loss": 2.8164,
|
| 955 |
+
"step": 134
|
| 956 |
+
},
|
| 957 |
+
{
|
| 958 |
+
"epoch": 0.007748397915793256,
|
| 959 |
+
"grad_norm": 0.2182077318429947,
|
| 960 |
+
"learning_rate": 0.0005543523825905229,
|
| 961 |
+
"loss": 2.8279,
|
| 962 |
+
"step": 135
|
| 963 |
+
},
|
| 964 |
+
{
|
| 965 |
+
"epoch": 0.007805793455910244,
|
| 966 |
+
"grad_norm": 0.24468722939491272,
|
| 967 |
+
"learning_rate": 0.0005536593973480297,
|
| 968 |
+
"loss": 2.8281,
|
| 969 |
+
"step": 136
|
| 970 |
+
},
|
| 971 |
+
{
|
| 972 |
+
"epoch": 0.007863188996027231,
|
| 973 |
+
"grad_norm": 0.22021321952342987,
|
| 974 |
+
"learning_rate": 0.0005529621510737175,
|
| 975 |
+
"loss": 2.8028,
|
| 976 |
+
"step": 137
|
| 977 |
+
},
|
| 978 |
+
{
|
| 979 |
+
"epoch": 0.007920584536144217,
|
| 980 |
+
"grad_norm": 0.20566654205322266,
|
| 981 |
+
"learning_rate": 0.0005522606724286498,
|
| 982 |
+
"loss": 2.7937,
|
| 983 |
+
"step": 138
|
| 984 |
+
},
|
| 985 |
+
{
|
| 986 |
+
"epoch": 0.007977980076261205,
|
| 987 |
+
"grad_norm": 0.1960543841123581,
|
| 988 |
+
"learning_rate": 0.0005515549902478665,
|
| 989 |
+
"loss": 2.8089,
|
| 990 |
+
"step": 139
|
| 991 |
+
},
|
| 992 |
+
{
|
| 993 |
+
"epoch": 0.008035375616378192,
|
| 994 |
+
"grad_norm": 0.2689999043941498,
|
| 995 |
+
"learning_rate": 0.0005508451335391975,
|
| 996 |
+
"loss": 2.7959,
|
| 997 |
+
"step": 140
|
| 998 |
+
},
|
| 999 |
+
{
|
| 1000 |
+
"epoch": 0.008092771156495178,
|
| 1001 |
+
"grad_norm": 0.19776718318462372,
|
| 1002 |
+
"learning_rate": 0.0005501311314820721,
|
| 1003 |
+
"loss": 2.8442,
|
| 1004 |
+
"step": 141
|
| 1005 |
+
},
|
| 1006 |
+
{
|
| 1007 |
+
"epoch": 0.008150166696612166,
|
| 1008 |
+
"grad_norm": 0.2156287282705307,
|
| 1009 |
+
"learning_rate": 0.0005494130134263184,
|
| 1010 |
+
"loss": 2.8224,
|
| 1011 |
+
"step": 142
|
| 1012 |
+
},
|
| 1013 |
+
{
|
| 1014 |
+
"epoch": 0.008207562236729153,
|
| 1015 |
+
"grad_norm": 0.17528703808784485,
|
| 1016 |
+
"learning_rate": 0.0005486908088909568,
|
| 1017 |
+
"loss": 2.8659,
|
| 1018 |
+
"step": 143
|
| 1019 |
+
},
|
| 1020 |
+
{
|
| 1021 |
+
"epoch": 0.00826495777684614,
|
| 1022 |
+
"grad_norm": 0.1757359504699707,
|
| 1023 |
+
"learning_rate": 0.0005479645475629872,
|
| 1024 |
+
"loss": 2.8119,
|
| 1025 |
+
"step": 144
|
| 1026 |
+
},
|
| 1027 |
+
{
|
| 1028 |
+
"epoch": 0.008322353316963127,
|
| 1029 |
+
"grad_norm": 0.1916513890028,
|
| 1030 |
+
"learning_rate": 0.0005472342592961683,
|
| 1031 |
+
"loss": 2.8069,
|
| 1032 |
+
"step": 145
|
| 1033 |
+
},
|
| 1034 |
+
{
|
| 1035 |
+
"epoch": 0.008379748857080115,
|
| 1036 |
+
"grad_norm": 0.19162799417972565,
|
| 1037 |
+
"learning_rate": 0.0005464999741097901,
|
| 1038 |
+
"loss": 2.8211,
|
| 1039 |
+
"step": 146
|
| 1040 |
+
},
|
| 1041 |
+
{
|
| 1042 |
+
"epoch": 0.0084371443971971,
|
| 1043 |
+
"grad_norm": 0.1881379634141922,
|
| 1044 |
+
"learning_rate": 0.0005457617221874408,
|
| 1045 |
+
"loss": 2.7954,
|
| 1046 |
+
"step": 147
|
| 1047 |
+
},
|
| 1048 |
+
{
|
| 1049 |
+
"epoch": 0.008494539937314088,
|
| 1050 |
+
"grad_norm": 0.22305060923099518,
|
| 1051 |
+
"learning_rate": 0.0005450195338757654,
|
| 1052 |
+
"loss": 2.8447,
|
| 1053 |
+
"step": 148
|
| 1054 |
+
},
|
| 1055 |
+
{
|
| 1056 |
+
"epoch": 0.008551935477431076,
|
| 1057 |
+
"grad_norm": 0.25081732869148254,
|
| 1058 |
+
"learning_rate": 0.0005442734396832185,
|
| 1059 |
+
"loss": 2.8205,
|
| 1060 |
+
"step": 149
|
| 1061 |
+
},
|
| 1062 |
+
{
|
| 1063 |
+
"epoch": 0.008609331017548063,
|
| 1064 |
+
"grad_norm": 0.24046167731285095,
|
| 1065 |
+
"learning_rate": 0.00054352347027881,
|
| 1066 |
+
"loss": 2.8246,
|
| 1067 |
+
"step": 150
|
| 1068 |
+
},
|
| 1069 |
+
{
|
| 1070 |
+
"epoch": 0.00866672655766505,
|
| 1071 |
+
"grad_norm": 0.20985569059848785,
|
| 1072 |
+
"learning_rate": 0.0005427696564908447,
|
| 1073 |
+
"loss": 2.8384,
|
| 1074 |
+
"step": 151
|
| 1075 |
+
},
|
| 1076 |
+
{
|
| 1077 |
+
"epoch": 0.008724122097782037,
|
| 1078 |
+
"grad_norm": 0.18979063630104065,
|
| 1079 |
+
"learning_rate": 0.000542012029305655,
|
| 1080 |
+
"loss": 2.8261,
|
| 1081 |
+
"step": 152
|
| 1082 |
+
},
|
| 1083 |
+
{
|
| 1084 |
+
"epoch": 0.008781517637899024,
|
| 1085 |
+
"grad_norm": 0.21513347327709198,
|
| 1086 |
+
"learning_rate": 0.0005412506198663268,
|
| 1087 |
+
"loss": 2.8197,
|
| 1088 |
+
"step": 153
|
| 1089 |
+
},
|
| 1090 |
+
{
|
| 1091 |
+
"epoch": 0.00883891317801601,
|
| 1092 |
+
"grad_norm": 0.25432831048965454,
|
| 1093 |
+
"learning_rate": 0.0005404854594714204,
|
| 1094 |
+
"loss": 2.8091,
|
| 1095 |
+
"step": 154
|
| 1096 |
+
},
|
| 1097 |
+
{
|
| 1098 |
+
"epoch": 0.008896308718132998,
|
| 1099 |
+
"grad_norm": 0.261273592710495,
|
| 1100 |
+
"learning_rate": 0.0005397165795736823,
|
| 1101 |
+
"loss": 2.8324,
|
| 1102 |
+
"step": 155
|
| 1103 |
+
},
|
| 1104 |
+
{
|
| 1105 |
+
"epoch": 0.008953704258249985,
|
| 1106 |
+
"grad_norm": 0.22144336998462677,
|
| 1107 |
+
"learning_rate": 0.0005389440117787538,
|
| 1108 |
+
"loss": 2.8459,
|
| 1109 |
+
"step": 156
|
| 1110 |
+
},
|
| 1111 |
+
{
|
| 1112 |
+
"epoch": 0.009011099798366971,
|
| 1113 |
+
"grad_norm": 0.1860560178756714,
|
| 1114 |
+
"learning_rate": 0.000538167787843871,
|
| 1115 |
+
"loss": 2.8552,
|
| 1116 |
+
"step": 157
|
| 1117 |
+
},
|
| 1118 |
+
{
|
| 1119 |
+
"epoch": 0.009068495338483959,
|
| 1120 |
+
"grad_norm": 0.2402401566505432,
|
| 1121 |
+
"learning_rate": 0.0005373879396765593,
|
| 1122 |
+
"loss": 2.8229,
|
| 1123 |
+
"step": 158
|
| 1124 |
+
},
|
| 1125 |
+
{
|
| 1126 |
+
"epoch": 0.009125890878600947,
|
| 1127 |
+
"grad_norm": 0.2112584114074707,
|
| 1128 |
+
"learning_rate": 0.0005366044993333228,
|
| 1129 |
+
"loss": 2.823,
|
| 1130 |
+
"step": 159
|
| 1131 |
+
},
|
| 1132 |
+
{
|
| 1133 |
+
"epoch": 0.009183286418717934,
|
| 1134 |
+
"grad_norm": 0.24757996201515198,
|
| 1135 |
+
"learning_rate": 0.0005358174990183254,
|
| 1136 |
+
"loss": 2.8458,
|
| 1137 |
+
"step": 160
|
| 1138 |
+
},
|
| 1139 |
+
{
|
| 1140 |
+
"epoch": 0.00924068195883492,
|
| 1141 |
+
"grad_norm": 0.20984984934329987,
|
| 1142 |
+
"learning_rate": 0.0005350269710820675,
|
| 1143 |
+
"loss": 2.8375,
|
| 1144 |
+
"step": 161
|
| 1145 |
+
},
|
| 1146 |
+
{
|
| 1147 |
+
"epoch": 0.009298077498951908,
|
| 1148 |
+
"grad_norm": 0.22329501807689667,
|
| 1149 |
+
"learning_rate": 0.0005342329480200562,
|
| 1150 |
+
"loss": 2.815,
|
| 1151 |
+
"step": 162
|
| 1152 |
+
},
|
| 1153 |
+
{
|
| 1154 |
+
"epoch": 0.009355473039068895,
|
| 1155 |
+
"grad_norm": 0.26144203543663025,
|
| 1156 |
+
"learning_rate": 0.0005334354624714697,
|
| 1157 |
+
"loss": 2.8286,
|
| 1158 |
+
"step": 163
|
| 1159 |
+
},
|
| 1160 |
+
{
|
| 1161 |
+
"epoch": 0.009412868579185881,
|
| 1162 |
+
"grad_norm": 0.20015327632427216,
|
| 1163 |
+
"learning_rate": 0.0005326345472178154,
|
| 1164 |
+
"loss": 2.8304,
|
| 1165 |
+
"step": 164
|
| 1166 |
+
},
|
| 1167 |
+
{
|
| 1168 |
+
"epoch": 0.009470264119302869,
|
| 1169 |
+
"grad_norm": 0.29256758093833923,
|
| 1170 |
+
"learning_rate": 0.0005318302351815823,
|
| 1171 |
+
"loss": 2.7884,
|
| 1172 |
+
"step": 165
|
| 1173 |
+
},
|
| 1174 |
+
{
|
| 1175 |
+
"epoch": 0.009527659659419856,
|
| 1176 |
+
"grad_norm": 0.22914084792137146,
|
| 1177 |
+
"learning_rate": 0.000531022559424888,
|
| 1178 |
+
"loss": 2.8253,
|
| 1179 |
+
"step": 166
|
| 1180 |
+
},
|
| 1181 |
+
{
|
| 1182 |
+
"epoch": 0.009585055199536842,
|
| 1183 |
+
"grad_norm": 0.2677003741264343,
|
| 1184 |
+
"learning_rate": 0.0005302115531481195,
|
| 1185 |
+
"loss": 2.8084,
|
| 1186 |
+
"step": 167
|
| 1187 |
+
},
|
| 1188 |
+
{
|
| 1189 |
+
"epoch": 0.00964245073965383,
|
| 1190 |
+
"grad_norm": 0.2672327756881714,
|
| 1191 |
+
"learning_rate": 0.000529397249688568,
|
| 1192 |
+
"loss": 2.8351,
|
| 1193 |
+
"step": 168
|
| 1194 |
+
},
|
| 1195 |
+
{
|
| 1196 |
+
"epoch": 0.009699846279770817,
|
| 1197 |
+
"grad_norm": 0.21281464397907257,
|
| 1198 |
+
"learning_rate": 0.0005285796825190598,
|
| 1199 |
+
"loss": 2.8463,
|
| 1200 |
+
"step": 169
|
| 1201 |
+
},
|
| 1202 |
+
{
|
| 1203 |
+
"epoch": 0.009757241819887805,
|
| 1204 |
+
"grad_norm": 0.22858156263828278,
|
| 1205 |
+
"learning_rate": 0.0005277588852465788,
|
| 1206 |
+
"loss": 2.8156,
|
| 1207 |
+
"step": 170
|
| 1208 |
+
},
|
| 1209 |
+
{
|
| 1210 |
+
"epoch": 0.009814637360004791,
|
| 1211 |
+
"grad_norm": 0.20694582164287567,
|
| 1212 |
+
"learning_rate": 0.0005269348916108859,
|
| 1213 |
+
"loss": 2.8392,
|
| 1214 |
+
"step": 171
|
| 1215 |
+
},
|
| 1216 |
+
{
|
| 1217 |
+
"epoch": 0.009872032900121779,
|
| 1218 |
+
"grad_norm": 0.22438685595989227,
|
| 1219 |
+
"learning_rate": 0.0005261077354831322,
|
| 1220 |
+
"loss": 2.8336,
|
| 1221 |
+
"step": 172
|
| 1222 |
+
},
|
| 1223 |
+
{
|
| 1224 |
+
"epoch": 0.009929428440238766,
|
| 1225 |
+
"grad_norm": 0.2279587984085083,
|
| 1226 |
+
"learning_rate": 0.0005252774508644666,
|
| 1227 |
+
"loss": 2.7972,
|
| 1228 |
+
"step": 173
|
| 1229 |
+
},
|
| 1230 |
+
{
|
| 1231 |
+
"epoch": 0.009986823980355752,
|
| 1232 |
+
"grad_norm": 0.21278439462184906,
|
| 1233 |
+
"learning_rate": 0.0005244440718846375,
|
| 1234 |
+
"loss": 2.7946,
|
| 1235 |
+
"step": 174
|
| 1236 |
+
},
|
| 1237 |
+
{
|
| 1238 |
+
"epoch": 0.01004421952047274,
|
| 1239 |
+
"grad_norm": 0.23399871587753296,
|
| 1240 |
+
"learning_rate": 0.0005236076328005906,
|
| 1241 |
+
"loss": 2.8648,
|
| 1242 |
+
"step": 175
|
| 1243 |
+
},
|
| 1244 |
+
{
|
| 1245 |
+
"epoch": 0.010101615060589727,
|
| 1246 |
+
"grad_norm": 0.2649572193622589,
|
| 1247 |
+
"learning_rate": 0.0005227681679950607,
|
| 1248 |
+
"loss": 2.8453,
|
| 1249 |
+
"step": 176
|
| 1250 |
+
},
|
| 1251 |
+
{
|
| 1252 |
+
"epoch": 0.010159010600706713,
|
| 1253 |
+
"grad_norm": 0.21067285537719727,
|
| 1254 |
+
"learning_rate": 0.0005219257119751581,
|
| 1255 |
+
"loss": 2.8357,
|
| 1256 |
+
"step": 177
|
| 1257 |
+
},
|
| 1258 |
+
{
|
| 1259 |
+
"epoch": 0.0102164061408237,
|
| 1260 |
+
"grad_norm": 0.22862860560417175,
|
| 1261 |
+
"learning_rate": 0.0005210802993709497,
|
| 1262 |
+
"loss": 2.8235,
|
| 1263 |
+
"step": 178
|
| 1264 |
+
},
|
| 1265 |
+
{
|
| 1266 |
+
"epoch": 0.010273801680940688,
|
| 1267 |
+
"grad_norm": 0.22179283201694489,
|
| 1268 |
+
"learning_rate": 0.0005202319649340369,
|
| 1269 |
+
"loss": 2.82,
|
| 1270 |
+
"step": 179
|
| 1271 |
+
},
|
| 1272 |
+
{
|
| 1273 |
+
"epoch": 0.010331197221057674,
|
| 1274 |
+
"grad_norm": 0.16690605878829956,
|
| 1275 |
+
"learning_rate": 0.0005193807435361252,
|
| 1276 |
+
"loss": 2.8237,
|
| 1277 |
+
"step": 180
|
| 1278 |
+
},
|
| 1279 |
+
{
|
| 1280 |
+
"epoch": 0.010388592761174662,
|
| 1281 |
+
"grad_norm": 0.21572506427764893,
|
| 1282 |
+
"learning_rate": 0.0005185266701675927,
|
| 1283 |
+
"loss": 2.8403,
|
| 1284 |
+
"step": 181
|
| 1285 |
+
},
|
| 1286 |
+
{
|
| 1287 |
+
"epoch": 0.01044598830129165,
|
| 1288 |
+
"grad_norm": 0.1778525710105896,
|
| 1289 |
+
"learning_rate": 0.0005176697799360502,
|
| 1290 |
+
"loss": 2.8204,
|
| 1291 |
+
"step": 182
|
| 1292 |
+
},
|
| 1293 |
+
{
|
| 1294 |
+
"epoch": 0.010503383841408637,
|
| 1295 |
+
"grad_norm": 0.18887534737586975,
|
| 1296 |
+
"learning_rate": 0.0005168101080648989,
|
| 1297 |
+
"loss": 2.8146,
|
| 1298 |
+
"step": 183
|
| 1299 |
+
},
|
| 1300 |
+
{
|
| 1301 |
+
"epoch": 0.010560779381525623,
|
| 1302 |
+
"grad_norm": 0.18108077347278595,
|
| 1303 |
+
"learning_rate": 0.0005159476898918823,
|
| 1304 |
+
"loss": 2.853,
|
| 1305 |
+
"step": 184
|
| 1306 |
+
},
|
| 1307 |
+
{
|
| 1308 |
+
"epoch": 0.01061817492164261,
|
| 1309 |
+
"grad_norm": 0.1870754212141037,
|
| 1310 |
+
"learning_rate": 0.0005150825608676336,
|
| 1311 |
+
"loss": 2.8537,
|
| 1312 |
+
"step": 185
|
| 1313 |
+
},
|
| 1314 |
+
{
|
| 1315 |
+
"epoch": 0.010675570461759598,
|
| 1316 |
+
"grad_norm": 0.16484060883522034,
|
| 1317 |
+
"learning_rate": 0.0005142147565542188,
|
| 1318 |
+
"loss": 2.8194,
|
| 1319 |
+
"step": 186
|
| 1320 |
+
},
|
| 1321 |
+
{
|
| 1322 |
+
"epoch": 0.010732966001876584,
|
| 1323 |
+
"grad_norm": 0.18527449667453766,
|
| 1324 |
+
"learning_rate": 0.0005133443126236739,
|
| 1325 |
+
"loss": 2.8402,
|
| 1326 |
+
"step": 187
|
| 1327 |
+
},
|
| 1328 |
+
{
|
| 1329 |
+
"epoch": 0.010790361541993572,
|
| 1330 |
+
"grad_norm": 0.17674389481544495,
|
| 1331 |
+
"learning_rate": 0.0005124712648565398,
|
| 1332 |
+
"loss": 2.8412,
|
| 1333 |
+
"step": 188
|
| 1334 |
+
},
|
| 1335 |
+
{
|
| 1336 |
+
"epoch": 0.01084775708211056,
|
| 1337 |
+
"grad_norm": 0.2521503269672394,
|
| 1338 |
+
"learning_rate": 0.0005115956491403907,
|
| 1339 |
+
"loss": 2.8348,
|
| 1340 |
+
"step": 189
|
| 1341 |
+
},
|
| 1342 |
+
{
|
| 1343 |
+
"epoch": 0.010905152622227545,
|
| 1344 |
+
"grad_norm": 0.17621657252311707,
|
| 1345 |
+
"learning_rate": 0.000510717501468359,
|
| 1346 |
+
"loss": 2.8293,
|
| 1347 |
+
"step": 190
|
| 1348 |
+
},
|
| 1349 |
+
{
|
| 1350 |
+
"epoch": 0.010962548162344533,
|
| 1351 |
+
"grad_norm": 0.2621336579322815,
|
| 1352 |
+
"learning_rate": 0.0005098368579376563,
|
| 1353 |
+
"loss": 2.8164,
|
| 1354 |
+
"step": 191
|
| 1355 |
+
},
|
| 1356 |
+
{
|
| 1357 |
+
"epoch": 0.01101994370246152,
|
| 1358 |
+
"grad_norm": 0.18950189650058746,
|
| 1359 |
+
"learning_rate": 0.0005089537547480885,
|
| 1360 |
+
"loss": 2.7976,
|
| 1361 |
+
"step": 192
|
| 1362 |
+
},
|
| 1363 |
+
{
|
| 1364 |
+
"epoch": 0.011077339242578508,
|
| 1365 |
+
"grad_norm": 0.24857239425182343,
|
| 1366 |
+
"learning_rate": 0.0005080682282005692,
|
| 1367 |
+
"loss": 2.8323,
|
| 1368 |
+
"step": 193
|
| 1369 |
+
},
|
| 1370 |
+
{
|
| 1371 |
+
"epoch": 0.011134734782695494,
|
| 1372 |
+
"grad_norm": 0.16708490252494812,
|
| 1373 |
+
"learning_rate": 0.0005071803146956262,
|
| 1374 |
+
"loss": 2.801,
|
| 1375 |
+
"step": 194
|
| 1376 |
+
},
|
| 1377 |
+
{
|
| 1378 |
+
"epoch": 0.011192130322812481,
|
| 1379 |
+
"grad_norm": 0.24443359673023224,
|
| 1380 |
+
"learning_rate": 0.000506290050731906,
|
| 1381 |
+
"loss": 2.8121,
|
| 1382 |
+
"step": 195
|
| 1383 |
+
},
|
| 1384 |
+
{
|
| 1385 |
+
"epoch": 0.011249525862929469,
|
| 1386 |
+
"grad_norm": 0.2458924949169159,
|
| 1387 |
+
"learning_rate": 0.0005053974729046734,
|
| 1388 |
+
"loss": 2.8325,
|
| 1389 |
+
"step": 196
|
| 1390 |
+
},
|
| 1391 |
+
{
|
| 1392 |
+
"epoch": 0.011306921403046455,
|
| 1393 |
+
"grad_norm": 0.2034812569618225,
|
| 1394 |
+
"learning_rate": 0.0005045026179043067,
|
| 1395 |
+
"loss": 2.8123,
|
| 1396 |
+
"step": 197
|
| 1397 |
+
},
|
| 1398 |
+
{
|
| 1399 |
+
"epoch": 0.011364316943163442,
|
| 1400 |
+
"grad_norm": 0.2774895429611206,
|
| 1401 |
+
"learning_rate": 0.0005036055225147901,
|
| 1402 |
+
"loss": 2.8324,
|
| 1403 |
+
"step": 198
|
| 1404 |
+
},
|
| 1405 |
+
{
|
| 1406 |
+
"epoch": 0.01142171248328043,
|
| 1407 |
+
"grad_norm": 0.22201013565063477,
|
| 1408 |
+
"learning_rate": 0.0005027062236122014,
|
| 1409 |
+
"loss": 2.8195,
|
| 1410 |
+
"step": 199
|
| 1411 |
+
},
|
| 1412 |
+
{
|
| 1413 |
+
"epoch": 0.011479108023397416,
|
| 1414 |
+
"grad_norm": 0.1893691122531891,
|
| 1415 |
+
"learning_rate": 0.0005018047581631961,
|
| 1416 |
+
"loss": 2.8177,
|
| 1417 |
+
"step": 200
|
| 1418 |
+
},
|
| 1419 |
+
{
|
| 1420 |
+
"epoch": 0.011479108023397416,
|
| 1421 |
+
"eval_loss": 2.749150037765503,
|
| 1422 |
+
"eval_runtime": 85.2258,
|
| 1423 |
+
"eval_samples_per_second": 50.63,
|
| 1424 |
+
"eval_steps_per_second": 12.66,
|
| 1425 |
+
"step": 200
|
| 1426 |
+
},
|
| 1427 |
+
{
|
| 1428 |
+
"epoch": 0.011536503563514404,
|
| 1429 |
+
"grad_norm": 0.2689765691757202,
|
| 1430 |
+
"learning_rate": 0.0005009011632234881,
|
| 1431 |
+
"loss": 2.8438,
|
| 1432 |
+
"step": 201
|
| 1433 |
+
},
|
| 1434 |
+
{
|
| 1435 |
+
"epoch": 0.011593899103631391,
|
| 1436 |
+
"grad_norm": 0.2234533727169037,
|
| 1437 |
+
"learning_rate": 0.0004999954759363262,
|
| 1438 |
+
"loss": 2.8103,
|
| 1439 |
+
"step": 202
|
| 1440 |
+
},
|
| 1441 |
+
{
|
| 1442 |
+
"epoch": 0.011651294643748379,
|
| 1443 |
+
"grad_norm": 0.25140801072120667,
|
| 1444 |
+
"learning_rate": 0.0004990877335309675,
|
| 1445 |
+
"loss": 2.8178,
|
| 1446 |
+
"step": 203
|
| 1447 |
+
},
|
| 1448 |
+
{
|
| 1449 |
+
"epoch": 0.011708690183865365,
|
| 1450 |
+
"grad_norm": 0.3070688843727112,
|
| 1451 |
+
"learning_rate": 0.0004981779733211468,
|
| 1452 |
+
"loss": 2.8518,
|
| 1453 |
+
"step": 204
|
| 1454 |
+
},
|
| 1455 |
+
{
|
| 1456 |
+
"epoch": 0.011766085723982352,
|
| 1457 |
+
"grad_norm": 0.25637757778167725,
|
| 1458 |
+
"learning_rate": 0.0004972662327035431,
|
| 1459 |
+
"loss": 2.8578,
|
| 1460 |
+
"step": 205
|
| 1461 |
+
},
|
| 1462 |
+
{
|
| 1463 |
+
"epoch": 0.01182348126409934,
|
| 1464 |
+
"grad_norm": 0.2551119923591614,
|
| 1465 |
+
"learning_rate": 0.0004963525491562421,
|
| 1466 |
+
"loss": 2.8237,
|
| 1467 |
+
"step": 206
|
| 1468 |
+
},
|
| 1469 |
+
{
|
| 1470 |
+
"epoch": 0.011880876804216326,
|
| 1471 |
+
"grad_norm": 0.2416735738515854,
|
| 1472 |
+
"learning_rate": 0.0004954369602371958,
|
| 1473 |
+
"loss": 2.8195,
|
| 1474 |
+
"step": 207
|
| 1475 |
+
},
|
| 1476 |
+
{
|
| 1477 |
+
"epoch": 0.011938272344333313,
|
| 1478 |
+
"grad_norm": 0.3950039744377136,
|
| 1479 |
+
"learning_rate": 0.0004945195035826785,
|
| 1480 |
+
"loss": 2.8087,
|
| 1481 |
+
"step": 208
|
| 1482 |
+
},
|
| 1483 |
+
{
|
| 1484 |
+
"epoch": 0.011995667884450301,
|
| 1485 |
+
"grad_norm": 0.16370531916618347,
|
| 1486 |
+
"learning_rate": 0.00049360021690574,
|
| 1487 |
+
"loss": 2.8464,
|
| 1488 |
+
"step": 209
|
| 1489 |
+
},
|
| 1490 |
+
{
|
| 1491 |
+
"epoch": 0.012053063424567287,
|
| 1492 |
+
"grad_norm": 0.28070008754730225,
|
| 1493 |
+
"learning_rate": 0.0004926791379946549,
|
| 1494 |
+
"loss": 2.8377,
|
| 1495 |
+
"step": 210
|
| 1496 |
+
},
|
| 1497 |
+
{
|
| 1498 |
+
"epoch": 0.012110458964684274,
|
| 1499 |
+
"grad_norm": 0.1902085244655609,
|
| 1500 |
+
"learning_rate": 0.0004917563047113695,
|
| 1501 |
+
"loss": 2.8279,
|
| 1502 |
+
"step": 211
|
| 1503 |
+
},
|
| 1504 |
+
{
|
| 1505 |
+
"epoch": 0.012167854504801262,
|
| 1506 |
+
"grad_norm": 0.27748385071754456,
|
| 1507 |
+
"learning_rate": 0.0004908317549899456,
|
| 1508 |
+
"loss": 2.837,
|
| 1509 |
+
"step": 212
|
| 1510 |
+
},
|
| 1511 |
+
{
|
| 1512 |
+
"epoch": 0.012225250044918248,
|
| 1513 |
+
"grad_norm": 0.18437190353870392,
|
| 1514 |
+
"learning_rate": 0.0004899055268350012,
|
| 1515 |
+
"loss": 2.8301,
|
| 1516 |
+
"step": 213
|
| 1517 |
+
},
|
| 1518 |
+
{
|
| 1519 |
+
"epoch": 0.012282645585035236,
|
| 1520 |
+
"grad_norm": 0.22971947491168976,
|
| 1521 |
+
"learning_rate": 0.0004889776583201479,
|
| 1522 |
+
"loss": 2.8051,
|
| 1523 |
+
"step": 214
|
| 1524 |
+
},
|
| 1525 |
+
{
|
| 1526 |
+
"epoch": 0.012340041125152223,
|
| 1527 |
+
"grad_norm": 0.238089457154274,
|
| 1528 |
+
"learning_rate": 0.0004880481875864261,
|
| 1529 |
+
"loss": 2.8162,
|
| 1530 |
+
"step": 215
|
| 1531 |
+
},
|
| 1532 |
+
{
|
| 1533 |
+
"epoch": 0.01239743666526921,
|
| 1534 |
+
"grad_norm": 0.24253320693969727,
|
| 1535 |
+
"learning_rate": 0.0004871171528407371,
|
| 1536 |
+
"loss": 2.8181,
|
| 1537 |
+
"step": 216
|
| 1538 |
+
},
|
| 1539 |
+
{
|
| 1540 |
+
"epoch": 0.012454832205386197,
|
| 1541 |
+
"grad_norm": 0.2351958006620407,
|
| 1542 |
+
"learning_rate": 0.0004861845923542728,
|
| 1543 |
+
"loss": 2.8136,
|
| 1544 |
+
"step": 217
|
| 1545 |
+
},
|
| 1546 |
+
{
|
| 1547 |
+
"epoch": 0.012512227745503184,
|
| 1548 |
+
"grad_norm": 0.23203608393669128,
|
| 1549 |
+
"learning_rate": 0.0004852505444609422,
|
| 1550 |
+
"loss": 2.804,
|
| 1551 |
+
"step": 218
|
| 1552 |
+
},
|
| 1553 |
+
{
|
| 1554 |
+
"epoch": 0.012569623285620172,
|
| 1555 |
+
"grad_norm": 0.1896822452545166,
|
| 1556 |
+
"learning_rate": 0.00048431504755579575,
|
| 1557 |
+
"loss": 2.8118,
|
| 1558 |
+
"step": 219
|
| 1559 |
+
},
|
| 1560 |
+
{
|
| 1561 |
+
"epoch": 0.012627018825737158,
|
| 1562 |
+
"grad_norm": 0.18357349932193756,
|
| 1563 |
+
"learning_rate": 0.0004833781400934471,
|
| 1564 |
+
"loss": 2.8205,
|
| 1565 |
+
"step": 220
|
| 1566 |
+
},
|
| 1567 |
+
{
|
| 1568 |
+
"epoch": 0.012684414365854145,
|
| 1569 |
+
"grad_norm": 0.23723295331001282,
|
| 1570 |
+
"learning_rate": 0.00048243986058649246,
|
| 1571 |
+
"loss": 2.8291,
|
| 1572 |
+
"step": 221
|
| 1573 |
+
},
|
| 1574 |
+
{
|
| 1575 |
+
"epoch": 0.012741809905971133,
|
| 1576 |
+
"grad_norm": 0.1937919706106186,
|
| 1577 |
+
"learning_rate": 0.0004815002476039273,
|
| 1578 |
+
"loss": 2.8416,
|
| 1579 |
+
"step": 222
|
| 1580 |
+
},
|
| 1581 |
+
{
|
| 1582 |
+
"epoch": 0.012799205446088119,
|
| 1583 |
+
"grad_norm": 0.19754467904567719,
|
| 1584 |
+
"learning_rate": 0.0004805593397695613,
|
| 1585 |
+
"loss": 2.7963,
|
| 1586 |
+
"step": 223
|
| 1587 |
+
},
|
| 1588 |
+
{
|
| 1589 |
+
"epoch": 0.012856600986205106,
|
| 1590 |
+
"grad_norm": 0.1592610776424408,
|
| 1591 |
+
"learning_rate": 0.00047961717576043,
|
| 1592 |
+
"loss": 2.8264,
|
| 1593 |
+
"step": 224
|
| 1594 |
+
},
|
| 1595 |
+
{
|
| 1596 |
+
"epoch": 0.012913996526322094,
|
| 1597 |
+
"grad_norm": 0.2083783745765686,
|
| 1598 |
+
"learning_rate": 0.00047867379430520585,
|
| 1599 |
+
"loss": 2.8348,
|
| 1600 |
+
"step": 225
|
| 1601 |
+
},
|
| 1602 |
+
{
|
| 1603 |
+
"epoch": 0.012971392066439082,
|
| 1604 |
+
"grad_norm": 0.1895647495985031,
|
| 1605 |
+
"learning_rate": 0.00047772923418260525,
|
| 1606 |
+
"loss": 2.8212,
|
| 1607 |
+
"step": 226
|
| 1608 |
+
},
|
| 1609 |
+
{
|
| 1610 |
+
"epoch": 0.013028787606556068,
|
| 1611 |
+
"grad_norm": 0.2173570841550827,
|
| 1612 |
+
"learning_rate": 0.0004767835342197954,
|
| 1613 |
+
"loss": 2.8098,
|
| 1614 |
+
"step": 227
|
| 1615 |
+
},
|
| 1616 |
+
{
|
| 1617 |
+
"epoch": 0.013086183146673055,
|
| 1618 |
+
"grad_norm": 0.1693475991487503,
|
| 1619 |
+
"learning_rate": 0.0004758367332907978,
|
| 1620 |
+
"loss": 2.796,
|
| 1621 |
+
"step": 228
|
| 1622 |
+
},
|
| 1623 |
+
{
|
| 1624 |
+
"epoch": 0.013143578686790043,
|
| 1625 |
+
"grad_norm": 0.21635355055332184,
|
| 1626 |
+
"learning_rate": 0.00047488887031489017,
|
| 1627 |
+
"loss": 2.843,
|
| 1628 |
+
"step": 229
|
| 1629 |
+
},
|
| 1630 |
+
{
|
| 1631 |
+
"epoch": 0.013200974226907029,
|
| 1632 |
+
"grad_norm": 0.18521156907081604,
|
| 1633 |
+
"learning_rate": 0.0004739399842550068,
|
| 1634 |
+
"loss": 2.8296,
|
| 1635 |
+
"step": 230
|
| 1636 |
+
},
|
| 1637 |
+
{
|
| 1638 |
+
"epoch": 0.013258369767024016,
|
| 1639 |
+
"grad_norm": 0.22925664484500885,
|
| 1640 |
+
"learning_rate": 0.00047299011411613734,
|
| 1641 |
+
"loss": 2.8287,
|
| 1642 |
+
"step": 231
|
| 1643 |
+
},
|
| 1644 |
+
{
|
| 1645 |
+
"epoch": 0.013315765307141004,
|
| 1646 |
+
"grad_norm": 0.24881386756896973,
|
| 1647 |
+
"learning_rate": 0.00047203929894372264,
|
| 1648 |
+
"loss": 2.8257,
|
| 1649 |
+
"step": 232
|
| 1650 |
+
},
|
| 1651 |
+
{
|
| 1652 |
+
"epoch": 0.01337316084725799,
|
| 1653 |
+
"grad_norm": 0.20801618695259094,
|
| 1654 |
+
"learning_rate": 0.00047108757782205043,
|
| 1655 |
+
"loss": 2.8241,
|
| 1656 |
+
"step": 233
|
| 1657 |
+
},
|
| 1658 |
+
{
|
| 1659 |
+
"epoch": 0.013430556387374977,
|
| 1660 |
+
"grad_norm": 0.199665367603302,
|
| 1661 |
+
"learning_rate": 0.0004701349898726483,
|
| 1662 |
+
"loss": 2.7916,
|
| 1663 |
+
"step": 234
|
| 1664 |
+
},
|
| 1665 |
+
{
|
| 1666 |
+
"epoch": 0.013487951927491965,
|
| 1667 |
+
"grad_norm": 0.25221607089042664,
|
| 1668 |
+
"learning_rate": 0.00046918157425267584,
|
| 1669 |
+
"loss": 2.8233,
|
| 1670 |
+
"step": 235
|
| 1671 |
+
},
|
| 1672 |
+
{
|
| 1673 |
+
"epoch": 0.013545347467608953,
|
| 1674 |
+
"grad_norm": 0.1931813657283783,
|
| 1675 |
+
"learning_rate": 0.00046822737015331505,
|
| 1676 |
+
"loss": 2.8016,
|
| 1677 |
+
"step": 236
|
| 1678 |
+
},
|
| 1679 |
+
{
|
| 1680 |
+
"epoch": 0.013602743007725938,
|
| 1681 |
+
"grad_norm": 0.17353369295597076,
|
| 1682 |
+
"learning_rate": 0.00046727241679815894,
|
| 1683 |
+
"loss": 2.8125,
|
| 1684 |
+
"step": 237
|
| 1685 |
+
},
|
| 1686 |
+
{
|
| 1687 |
+
"epoch": 0.013660138547842926,
|
| 1688 |
+
"grad_norm": 0.22225958108901978,
|
| 1689 |
+
"learning_rate": 0.0004663167534415996,
|
| 1690 |
+
"loss": 2.824,
|
| 1691 |
+
"step": 238
|
| 1692 |
+
},
|
| 1693 |
+
{
|
| 1694 |
+
"epoch": 0.013717534087959914,
|
| 1695 |
+
"grad_norm": 0.17010116577148438,
|
| 1696 |
+
"learning_rate": 0.0004653604193672147,
|
| 1697 |
+
"loss": 2.8425,
|
| 1698 |
+
"step": 239
|
| 1699 |
+
},
|
| 1700 |
+
{
|
| 1701 |
+
"epoch": 0.0137749296280769,
|
| 1702 |
+
"grad_norm": 0.2103683203458786,
|
| 1703 |
+
"learning_rate": 0.00046440345388615225,
|
| 1704 |
+
"loss": 2.8641,
|
| 1705 |
+
"step": 240
|
| 1706 |
+
},
|
| 1707 |
+
{
|
| 1708 |
+
"epoch": 0.013832325168193887,
|
| 1709 |
+
"grad_norm": 0.17934557795524597,
|
| 1710 |
+
"learning_rate": 0.00046344589633551497,
|
| 1711 |
+
"loss": 2.8069,
|
| 1712 |
+
"step": 241
|
| 1713 |
+
},
|
| 1714 |
+
{
|
| 1715 |
+
"epoch": 0.013889720708310875,
|
| 1716 |
+
"grad_norm": 0.2116999328136444,
|
| 1717 |
+
"learning_rate": 0.0004624877860767434,
|
| 1718 |
+
"loss": 2.8601,
|
| 1719 |
+
"step": 242
|
| 1720 |
+
},
|
| 1721 |
+
{
|
| 1722 |
+
"epoch": 0.01394711624842786,
|
| 1723 |
+
"grad_norm": 0.20861205458641052,
|
| 1724 |
+
"learning_rate": 0.0004615291624939975,
|
| 1725 |
+
"loss": 2.8232,
|
| 1726 |
+
"step": 243
|
| 1727 |
+
},
|
| 1728 |
+
{
|
| 1729 |
+
"epoch": 0.014004511788544848,
|
| 1730 |
+
"grad_norm": 0.24393285810947418,
|
| 1731 |
+
"learning_rate": 0.0004605700649925381,
|
| 1732 |
+
"loss": 2.8041,
|
| 1733 |
+
"step": 244
|
| 1734 |
+
},
|
| 1735 |
+
{
|
| 1736 |
+
"epoch": 0.014061907328661836,
|
| 1737 |
+
"grad_norm": 0.2089577168226242,
|
| 1738 |
+
"learning_rate": 0.0004596105329971069,
|
| 1739 |
+
"loss": 2.8351,
|
| 1740 |
+
"step": 245
|
| 1741 |
+
},
|
| 1742 |
+
{
|
| 1743 |
+
"epoch": 0.014119302868778822,
|
| 1744 |
+
"grad_norm": 0.20232421159744263,
|
| 1745 |
+
"learning_rate": 0.00045865060595030616,
|
| 1746 |
+
"loss": 2.8171,
|
| 1747 |
+
"step": 246
|
| 1748 |
+
},
|
| 1749 |
+
{
|
| 1750 |
+
"epoch": 0.01417669840889581,
|
| 1751 |
+
"grad_norm": 0.22081732749938965,
|
| 1752 |
+
"learning_rate": 0.00045769032331097686,
|
| 1753 |
+
"loss": 2.8202,
|
| 1754 |
+
"step": 247
|
| 1755 |
+
},
|
| 1756 |
+
{
|
| 1757 |
+
"epoch": 0.014234093949012797,
|
| 1758 |
+
"grad_norm": 0.17081516981124878,
|
| 1759 |
+
"learning_rate": 0.00045672972455257723,
|
| 1760 |
+
"loss": 2.8358,
|
| 1761 |
+
"step": 248
|
| 1762 |
+
},
|
| 1763 |
+
{
|
| 1764 |
+
"epoch": 0.014291489489129785,
|
| 1765 |
+
"grad_norm": 0.3317008316516876,
|
| 1766 |
+
"learning_rate": 0.0004557688491615597,
|
| 1767 |
+
"loss": 2.8302,
|
| 1768 |
+
"step": 249
|
| 1769 |
+
},
|
| 1770 |
+
{
|
| 1771 |
+
"epoch": 0.01434888502924677,
|
| 1772 |
+
"grad_norm": 0.23239760100841522,
|
| 1773 |
+
"learning_rate": 0.0004548077366357483,
|
| 1774 |
+
"loss": 2.8191,
|
| 1775 |
+
"step": 250
|
| 1776 |
+
},
|
| 1777 |
+
{
|
| 1778 |
+
"epoch": 0.014406280569363758,
|
| 1779 |
+
"grad_norm": 0.22138993442058563,
|
| 1780 |
+
"learning_rate": 0.0004538464264827143,
|
| 1781 |
+
"loss": 2.8096,
|
| 1782 |
+
"step": 251
|
| 1783 |
+
},
|
| 1784 |
+
{
|
| 1785 |
+
"epoch": 0.014463676109480746,
|
| 1786 |
+
"grad_norm": 0.23655574023723602,
|
| 1787 |
+
"learning_rate": 0.000452884958218153,
|
| 1788 |
+
"loss": 2.8295,
|
| 1789 |
+
"step": 252
|
| 1790 |
+
},
|
| 1791 |
+
{
|
| 1792 |
+
"epoch": 0.014521071649597731,
|
| 1793 |
+
"grad_norm": 0.2227945327758789,
|
| 1794 |
+
"learning_rate": 0.000451923371364259,
|
| 1795 |
+
"loss": 2.8158,
|
| 1796 |
+
"step": 253
|
| 1797 |
+
},
|
| 1798 |
+
{
|
| 1799 |
+
"epoch": 0.014578467189714719,
|
| 1800 |
+
"grad_norm": 0.20443300902843475,
|
| 1801 |
+
"learning_rate": 0.0004509617054481017,
|
| 1802 |
+
"loss": 2.83,
|
| 1803 |
+
"step": 254
|
| 1804 |
+
},
|
| 1805 |
+
{
|
| 1806 |
+
"epoch": 0.014635862729831707,
|
| 1807 |
+
"grad_norm": 0.22221451997756958,
|
| 1808 |
+
"learning_rate": 0.00045,
|
| 1809 |
+
"loss": 2.8253,
|
| 1810 |
+
"step": 255
|
| 1811 |
+
},
|
| 1812 |
+
{
|
| 1813 |
+
"epoch": 0.014693258269948693,
|
| 1814 |
+
"grad_norm": 0.1941068023443222,
|
| 1815 |
+
"learning_rate": 0.00044903829455189825,
|
| 1816 |
+
"loss": 2.83,
|
| 1817 |
+
"step": 256
|
| 1818 |
+
},
|
| 1819 |
+
{
|
| 1820 |
+
"epoch": 0.01475065381006568,
|
| 1821 |
+
"grad_norm": 0.1914331614971161,
|
| 1822 |
+
"learning_rate": 0.0004480766286357409,
|
| 1823 |
+
"loss": 2.8162,
|
| 1824 |
+
"step": 257
|
| 1825 |
+
},
|
| 1826 |
+
{
|
| 1827 |
+
"epoch": 0.014808049350182668,
|
| 1828 |
+
"grad_norm": 0.21014779806137085,
|
| 1829 |
+
"learning_rate": 0.0004471150417818469,
|
| 1830 |
+
"loss": 2.7993,
|
| 1831 |
+
"step": 258
|
| 1832 |
+
},
|
| 1833 |
+
{
|
| 1834 |
+
"epoch": 0.014865444890299655,
|
| 1835 |
+
"grad_norm": 0.2057676762342453,
|
| 1836 |
+
"learning_rate": 0.00044615357351728566,
|
| 1837 |
+
"loss": 2.8223,
|
| 1838 |
+
"step": 259
|
| 1839 |
+
},
|
| 1840 |
+
{
|
| 1841 |
+
"epoch": 0.014922840430416641,
|
| 1842 |
+
"grad_norm": 0.19875939190387726,
|
| 1843 |
+
"learning_rate": 0.00044519226336425165,
|
| 1844 |
+
"loss": 2.8016,
|
| 1845 |
+
"step": 260
|
| 1846 |
+
},
|
| 1847 |
+
{
|
| 1848 |
+
"epoch": 0.014980235970533629,
|
| 1849 |
+
"grad_norm": 0.23691999912261963,
|
| 1850 |
+
"learning_rate": 0.0004442311508384402,
|
| 1851 |
+
"loss": 2.8373,
|
| 1852 |
+
"step": 261
|
| 1853 |
+
},
|
| 1854 |
+
{
|
| 1855 |
+
"epoch": 0.015037631510650616,
|
| 1856 |
+
"grad_norm": 0.1729947328567505,
|
| 1857 |
+
"learning_rate": 0.0004432702754474228,
|
| 1858 |
+
"loss": 2.8233,
|
| 1859 |
+
"step": 262
|
| 1860 |
+
},
|
| 1861 |
+
{
|
| 1862 |
+
"epoch": 0.015095027050767602,
|
| 1863 |
+
"grad_norm": 0.18821187317371368,
|
| 1864 |
+
"learning_rate": 0.00044230967668902306,
|
| 1865 |
+
"loss": 2.8128,
|
| 1866 |
+
"step": 263
|
| 1867 |
+
},
|
| 1868 |
+
{
|
| 1869 |
+
"epoch": 0.01515242259088459,
|
| 1870 |
+
"grad_norm": 0.2283882200717926,
|
| 1871 |
+
"learning_rate": 0.00044134939404969387,
|
| 1872 |
+
"loss": 2.8178,
|
| 1873 |
+
"step": 264
|
| 1874 |
+
},
|
| 1875 |
+
{
|
| 1876 |
+
"epoch": 0.015209818131001578,
|
| 1877 |
+
"grad_norm": 0.16724412143230438,
|
| 1878 |
+
"learning_rate": 0.000440389467002893,
|
| 1879 |
+
"loss": 2.8249,
|
| 1880 |
+
"step": 265
|
| 1881 |
+
},
|
| 1882 |
+
{
|
| 1883 |
+
"epoch": 0.015267213671118563,
|
| 1884 |
+
"grad_norm": 0.18209712207317352,
|
| 1885 |
+
"learning_rate": 0.00043942993500746183,
|
| 1886 |
+
"loss": 2.8095,
|
| 1887 |
+
"step": 266
|
| 1888 |
+
},
|
| 1889 |
+
{
|
| 1890 |
+
"epoch": 0.015324609211235551,
|
| 1891 |
+
"grad_norm": 0.1857995092868805,
|
| 1892 |
+
"learning_rate": 0.00043847083750600253,
|
| 1893 |
+
"loss": 2.806,
|
| 1894 |
+
"step": 267
|
| 1895 |
+
},
|
| 1896 |
+
{
|
| 1897 |
+
"epoch": 0.015382004751352539,
|
| 1898 |
+
"grad_norm": 0.20734605193138123,
|
| 1899 |
+
"learning_rate": 0.0004375122139232566,
|
| 1900 |
+
"loss": 2.8695,
|
| 1901 |
+
"step": 268
|
| 1902 |
+
},
|
| 1903 |
+
{
|
| 1904 |
+
"epoch": 0.015439400291469526,
|
| 1905 |
+
"grad_norm": 0.23138895630836487,
|
| 1906 |
+
"learning_rate": 0.00043655410366448495,
|
| 1907 |
+
"loss": 2.8033,
|
| 1908 |
+
"step": 269
|
| 1909 |
+
},
|
| 1910 |
+
{
|
| 1911 |
+
"epoch": 0.015496795831586512,
|
| 1912 |
+
"grad_norm": 0.20481987297534943,
|
| 1913 |
+
"learning_rate": 0.0004355965461138477,
|
| 1914 |
+
"loss": 2.8269,
|
| 1915 |
+
"step": 270
|
| 1916 |
+
},
|
| 1917 |
+
{
|
| 1918 |
+
"epoch": 0.0155541913717035,
|
| 1919 |
+
"grad_norm": 0.2318529337644577,
|
| 1920 |
+
"learning_rate": 0.00043463958063278524,
|
| 1921 |
+
"loss": 2.8332,
|
| 1922 |
+
"step": 271
|
| 1923 |
+
},
|
| 1924 |
+
{
|
| 1925 |
+
"epoch": 0.015611586911820487,
|
| 1926 |
+
"grad_norm": 0.2501411736011505,
|
| 1927 |
+
"learning_rate": 0.00043368324655840035,
|
| 1928 |
+
"loss": 2.8445,
|
| 1929 |
+
"step": 272
|
| 1930 |
+
},
|
| 1931 |
+
{
|
| 1932 |
+
"epoch": 0.015668982451937475,
|
| 1933 |
+
"grad_norm": 0.26137158274650574,
|
| 1934 |
+
"learning_rate": 0.0004327275832018411,
|
| 1935 |
+
"loss": 2.8279,
|
| 1936 |
+
"step": 273
|
| 1937 |
+
},
|
| 1938 |
+
{
|
| 1939 |
+
"epoch": 0.015726377992054463,
|
| 1940 |
+
"grad_norm": 0.19074887037277222,
|
| 1941 |
+
"learning_rate": 0.0004317726298466849,
|
| 1942 |
+
"loss": 2.8132,
|
| 1943 |
+
"step": 274
|
| 1944 |
+
},
|
| 1945 |
+
{
|
| 1946 |
+
"epoch": 0.015783773532171447,
|
| 1947 |
+
"grad_norm": 0.26000818610191345,
|
| 1948 |
+
"learning_rate": 0.0004308184257473241,
|
| 1949 |
+
"loss": 2.8091,
|
| 1950 |
+
"step": 275
|
| 1951 |
+
},
|
| 1952 |
+
{
|
| 1953 |
+
"epoch": 0.015841169072288434,
|
| 1954 |
+
"grad_norm": 0.16060984134674072,
|
| 1955 |
+
"learning_rate": 0.0004298650101273517,
|
| 1956 |
+
"loss": 2.8206,
|
| 1957 |
+
"step": 276
|
| 1958 |
+
},
|
| 1959 |
+
{
|
| 1960 |
+
"epoch": 0.015898564612405422,
|
| 1961 |
+
"grad_norm": 0.284445583820343,
|
| 1962 |
+
"learning_rate": 0.00042891242217794954,
|
| 1963 |
+
"loss": 2.7867,
|
| 1964 |
+
"step": 277
|
| 1965 |
+
},
|
| 1966 |
+
{
|
| 1967 |
+
"epoch": 0.01595596015252241,
|
| 1968 |
+
"grad_norm": 0.15903466939926147,
|
| 1969 |
+
"learning_rate": 0.0004279607010562773,
|
| 1970 |
+
"loss": 2.83,
|
| 1971 |
+
"step": 278
|
| 1972 |
+
},
|
| 1973 |
+
{
|
| 1974 |
+
"epoch": 0.016013355692639397,
|
| 1975 |
+
"grad_norm": 0.24330751597881317,
|
| 1976 |
+
"learning_rate": 0.0004270098858838626,
|
| 1977 |
+
"loss": 2.817,
|
| 1978 |
+
"step": 279
|
| 1979 |
+
},
|
| 1980 |
+
{
|
| 1981 |
+
"epoch": 0.016070751232756385,
|
| 1982 |
+
"grad_norm": 0.1687777042388916,
|
| 1983 |
+
"learning_rate": 0.0004260600157449931,
|
| 1984 |
+
"loss": 2.8112,
|
| 1985 |
+
"step": 280
|
| 1986 |
+
},
|
| 1987 |
+
{
|
| 1988 |
+
"epoch": 0.01612814677287337,
|
| 1989 |
+
"grad_norm": 0.18230785429477692,
|
| 1990 |
+
"learning_rate": 0.0004251111296851098,
|
| 1991 |
+
"loss": 2.8394,
|
| 1992 |
+
"step": 281
|
| 1993 |
+
},
|
| 1994 |
+
{
|
| 1995 |
+
"epoch": 0.016185542312990357,
|
| 1996 |
+
"grad_norm": 0.1889660507440567,
|
| 1997 |
+
"learning_rate": 0.00042416326670920217,
|
| 1998 |
+
"loss": 2.8109,
|
| 1999 |
+
"step": 282
|
| 2000 |
+
},
|
| 2001 |
+
{
|
| 2002 |
+
"epoch": 0.016242937853107344,
|
| 2003 |
+
"grad_norm": 0.16135123372077942,
|
| 2004 |
+
"learning_rate": 0.0004232164657802045,
|
| 2005 |
+
"loss": 2.7953,
|
| 2006 |
+
"step": 283
|
| 2007 |
+
},
|
| 2008 |
+
{
|
| 2009 |
+
"epoch": 0.016300333393224332,
|
| 2010 |
+
"grad_norm": 0.15787218511104584,
|
| 2011 |
+
"learning_rate": 0.00042227076581739467,
|
| 2012 |
+
"loss": 2.7921,
|
| 2013 |
+
"step": 284
|
| 2014 |
+
},
|
| 2015 |
+
{
|
| 2016 |
+
"epoch": 0.01635772893334132,
|
| 2017 |
+
"grad_norm": 0.16313977539539337,
|
| 2018 |
+
"learning_rate": 0.0004213262056947942,
|
| 2019 |
+
"loss": 2.8107,
|
| 2020 |
+
"step": 285
|
| 2021 |
+
},
|
| 2022 |
+
{
|
| 2023 |
+
"epoch": 0.016415124473458307,
|
| 2024 |
+
"grad_norm": 0.18806132674217224,
|
| 2025 |
+
"learning_rate": 0.0004203828242395699,
|
| 2026 |
+
"loss": 2.8451,
|
| 2027 |
+
"step": 286
|
| 2028 |
+
},
|
| 2029 |
+
{
|
| 2030 |
+
"epoch": 0.016472520013575295,
|
| 2031 |
+
"grad_norm": 0.17279674112796783,
|
| 2032 |
+
"learning_rate": 0.00041944066023043866,
|
| 2033 |
+
"loss": 2.8333,
|
| 2034 |
+
"step": 287
|
| 2035 |
+
},
|
| 2036 |
+
{
|
| 2037 |
+
"epoch": 0.01652991555369228,
|
| 2038 |
+
"grad_norm": 0.17451834678649902,
|
| 2039 |
+
"learning_rate": 0.00041849975239607255,
|
| 2040 |
+
"loss": 2.7798,
|
| 2041 |
+
"step": 288
|
| 2042 |
+
},
|
| 2043 |
+
{
|
| 2044 |
+
"epoch": 0.016587311093809266,
|
| 2045 |
+
"grad_norm": 0.1943039745092392,
|
| 2046 |
+
"learning_rate": 0.00041756013941350747,
|
| 2047 |
+
"loss": 2.8011,
|
| 2048 |
+
"step": 289
|
| 2049 |
+
},
|
| 2050 |
+
{
|
| 2051 |
+
"epoch": 0.016644706633926254,
|
| 2052 |
+
"grad_norm": 0.1578904092311859,
|
| 2053 |
+
"learning_rate": 0.0004166218599065528,
|
| 2054 |
+
"loss": 2.852,
|
| 2055 |
+
"step": 290
|
| 2056 |
+
},
|
| 2057 |
+
{
|
| 2058 |
+
"epoch": 0.01670210217404324,
|
| 2059 |
+
"grad_norm": 0.20066620409488678,
|
| 2060 |
+
"learning_rate": 0.0004156849524442042,
|
| 2061 |
+
"loss": 2.7876,
|
| 2062 |
+
"step": 291
|
| 2063 |
+
},
|
| 2064 |
+
{
|
| 2065 |
+
"epoch": 0.01675949771416023,
|
| 2066 |
+
"grad_norm": 0.18306495249271393,
|
| 2067 |
+
"learning_rate": 0.0004147494555390577,
|
| 2068 |
+
"loss": 2.817,
|
| 2069 |
+
"step": 292
|
| 2070 |
+
},
|
| 2071 |
+
{
|
| 2072 |
+
"epoch": 0.016816893254277217,
|
| 2073 |
+
"grad_norm": 0.1622687727212906,
|
| 2074 |
+
"learning_rate": 0.0004138154076457271,
|
| 2075 |
+
"loss": 2.815,
|
| 2076 |
+
"step": 293
|
| 2077 |
+
},
|
| 2078 |
+
{
|
| 2079 |
+
"epoch": 0.0168742887943942,
|
| 2080 |
+
"grad_norm": 0.2056518942117691,
|
| 2081 |
+
"learning_rate": 0.0004128828471592628,
|
| 2082 |
+
"loss": 2.8131,
|
| 2083 |
+
"step": 294
|
| 2084 |
+
},
|
| 2085 |
+
{
|
| 2086 |
+
"epoch": 0.01693168433451119,
|
| 2087 |
+
"grad_norm": 0.17123937606811523,
|
| 2088 |
+
"learning_rate": 0.00041195181241357383,
|
| 2089 |
+
"loss": 2.8025,
|
| 2090 |
+
"step": 295
|
| 2091 |
+
},
|
| 2092 |
+
{
|
| 2093 |
+
"epoch": 0.016989079874628176,
|
| 2094 |
+
"grad_norm": 0.2233334332704544,
|
| 2095 |
+
"learning_rate": 0.00041102234167985204,
|
| 2096 |
+
"loss": 2.8347,
|
| 2097 |
+
"step": 296
|
| 2098 |
+
},
|
| 2099 |
+
{
|
| 2100 |
+
"epoch": 0.017046475414745164,
|
| 2101 |
+
"grad_norm": 0.20740529894828796,
|
| 2102 |
+
"learning_rate": 0.0004100944731649987,
|
| 2103 |
+
"loss": 2.8099,
|
| 2104 |
+
"step": 297
|
| 2105 |
+
},
|
| 2106 |
+
{
|
| 2107 |
+
"epoch": 0.01710387095486215,
|
| 2108 |
+
"grad_norm": 0.20391066372394562,
|
| 2109 |
+
"learning_rate": 0.0004091682450100543,
|
| 2110 |
+
"loss": 2.8363,
|
| 2111 |
+
"step": 298
|
| 2112 |
+
},
|
| 2113 |
+
{
|
| 2114 |
+
"epoch": 0.01716126649497914,
|
| 2115 |
+
"grad_norm": 0.17306548357009888,
|
| 2116 |
+
"learning_rate": 0.0004082436952886305,
|
| 2117 |
+
"loss": 2.8211,
|
| 2118 |
+
"step": 299
|
| 2119 |
+
},
|
| 2120 |
+
{
|
| 2121 |
+
"epoch": 0.017218662035096127,
|
| 2122 |
+
"grad_norm": 0.24933576583862305,
|
| 2123 |
+
"learning_rate": 0.0004073208620053451,
|
| 2124 |
+
"loss": 2.8048,
|
| 2125 |
+
"step": 300
|
| 2126 |
+
},
|
| 2127 |
+
{
|
| 2128 |
+
"epoch": 0.017218662035096127,
|
| 2129 |
+
"eval_loss": 2.7432332038879395,
|
| 2130 |
+
"eval_runtime": 85.2508,
|
| 2131 |
+
"eval_samples_per_second": 50.615,
|
| 2132 |
+
"eval_steps_per_second": 12.657,
|
| 2133 |
+
"step": 300
|
| 2134 |
+
},
|
| 2135 |
+
{
|
| 2136 |
+
"epoch": 0.01727605757521311,
|
| 2137 |
+
"grad_norm": 0.231708824634552,
|
| 2138 |
+
"learning_rate": 0.00040639978309425995,
|
| 2139 |
+
"loss": 2.8025,
|
| 2140 |
+
"step": 301
|
| 2141 |
+
},
|
| 2142 |
+
{
|
| 2143 |
+
"epoch": 0.0173334531153301,
|
| 2144 |
+
"grad_norm": 0.15970614552497864,
|
| 2145 |
+
"learning_rate": 0.00040548049641732137,
|
| 2146 |
+
"loss": 2.8392,
|
| 2147 |
+
"step": 302
|
| 2148 |
+
},
|
| 2149 |
+
{
|
| 2150 |
+
"epoch": 0.017390848655447086,
|
| 2151 |
+
"grad_norm": 0.20457029342651367,
|
| 2152 |
+
"learning_rate": 0.0004045630397628042,
|
| 2153 |
+
"loss": 2.8247,
|
| 2154 |
+
"step": 303
|
| 2155 |
+
},
|
| 2156 |
+
{
|
| 2157 |
+
"epoch": 0.017448244195564074,
|
| 2158 |
+
"grad_norm": 0.1734900325536728,
|
| 2159 |
+
"learning_rate": 0.00040364745084375787,
|
| 2160 |
+
"loss": 2.7979,
|
| 2161 |
+
"step": 304
|
| 2162 |
+
},
|
| 2163 |
+
{
|
| 2164 |
+
"epoch": 0.01750563973568106,
|
| 2165 |
+
"grad_norm": 0.19265452027320862,
|
| 2166 |
+
"learning_rate": 0.00040273376729645685,
|
| 2167 |
+
"loss": 2.8033,
|
| 2168 |
+
"step": 305
|
| 2169 |
+
},
|
| 2170 |
+
{
|
| 2171 |
+
"epoch": 0.01756303527579805,
|
| 2172 |
+
"grad_norm": 0.19174844026565552,
|
| 2173 |
+
"learning_rate": 0.00040182202667885317,
|
| 2174 |
+
"loss": 2.8354,
|
| 2175 |
+
"step": 306
|
| 2176 |
+
},
|
| 2177 |
+
{
|
| 2178 |
+
"epoch": 0.017620430815915036,
|
| 2179 |
+
"grad_norm": 0.27793413400650024,
|
| 2180 |
+
"learning_rate": 0.00040091226646903245,
|
| 2181 |
+
"loss": 2.797,
|
| 2182 |
+
"step": 307
|
| 2183 |
+
},
|
| 2184 |
+
{
|
| 2185 |
+
"epoch": 0.01767782635603202,
|
| 2186 |
+
"grad_norm": 0.1806309074163437,
|
| 2187 |
+
"learning_rate": 0.00040000452406367367,
|
| 2188 |
+
"loss": 2.8046,
|
| 2189 |
+
"step": 308
|
| 2190 |
+
},
|
| 2191 |
+
{
|
| 2192 |
+
"epoch": 0.017735221896149008,
|
| 2193 |
+
"grad_norm": 0.2249089479446411,
|
| 2194 |
+
"learning_rate": 0.0003990988367765118,
|
| 2195 |
+
"loss": 2.8125,
|
| 2196 |
+
"step": 309
|
| 2197 |
+
},
|
| 2198 |
+
{
|
| 2199 |
+
"epoch": 0.017792617436265996,
|
| 2200 |
+
"grad_norm": 0.27839699387550354,
|
| 2201 |
+
"learning_rate": 0.00039819524183680384,
|
| 2202 |
+
"loss": 2.8183,
|
| 2203 |
+
"step": 310
|
| 2204 |
+
},
|
| 2205 |
+
{
|
| 2206 |
+
"epoch": 0.017850012976382983,
|
| 2207 |
+
"grad_norm": 0.1877232789993286,
|
| 2208 |
+
"learning_rate": 0.00039729377638779857,
|
| 2209 |
+
"loss": 2.7989,
|
| 2210 |
+
"step": 311
|
| 2211 |
+
},
|
| 2212 |
+
{
|
| 2213 |
+
"epoch": 0.01790740851649997,
|
| 2214 |
+
"grad_norm": 0.25160273909568787,
|
| 2215 |
+
"learning_rate": 0.00039639447748520985,
|
| 2216 |
+
"loss": 2.8536,
|
| 2217 |
+
"step": 312
|
| 2218 |
+
},
|
| 2219 |
+
{
|
| 2220 |
+
"epoch": 0.01796480405661696,
|
| 2221 |
+
"grad_norm": 0.23843353986740112,
|
| 2222 |
+
"learning_rate": 0.0003954973820956932,
|
| 2223 |
+
"loss": 2.8064,
|
| 2224 |
+
"step": 313
|
| 2225 |
+
},
|
| 2226 |
+
{
|
| 2227 |
+
"epoch": 0.018022199596733943,
|
| 2228 |
+
"grad_norm": 0.2549470365047455,
|
| 2229 |
+
"learning_rate": 0.00039460252709532656,
|
| 2230 |
+
"loss": 2.8415,
|
| 2231 |
+
"step": 314
|
| 2232 |
+
},
|
| 2233 |
+
{
|
| 2234 |
+
"epoch": 0.01807959513685093,
|
| 2235 |
+
"grad_norm": 0.39248892664909363,
|
| 2236 |
+
"learning_rate": 0.0003937099492680938,
|
| 2237 |
+
"loss": 2.8137,
|
| 2238 |
+
"step": 315
|
| 2239 |
+
},
|
| 2240 |
+
{
|
| 2241 |
+
"epoch": 0.018136990676967918,
|
| 2242 |
+
"grad_norm": 0.24034982919692993,
|
| 2243 |
+
"learning_rate": 0.0003928196853043737,
|
| 2244 |
+
"loss": 2.8301,
|
| 2245 |
+
"step": 316
|
| 2246 |
+
},
|
| 2247 |
+
{
|
| 2248 |
+
"epoch": 0.018194386217084905,
|
| 2249 |
+
"grad_norm": 0.29434794187545776,
|
| 2250 |
+
"learning_rate": 0.00039193177179943083,
|
| 2251 |
+
"loss": 2.8288,
|
| 2252 |
+
"step": 317
|
| 2253 |
+
},
|
| 2254 |
+
{
|
| 2255 |
+
"epoch": 0.018251781757201893,
|
| 2256 |
+
"grad_norm": 0.21636317670345306,
|
| 2257 |
+
"learning_rate": 0.0003910462452519114,
|
| 2258 |
+
"loss": 2.8121,
|
| 2259 |
+
"step": 318
|
| 2260 |
+
},
|
| 2261 |
+
{
|
| 2262 |
+
"epoch": 0.01830917729731888,
|
| 2263 |
+
"grad_norm": 0.2217407375574112,
|
| 2264 |
+
"learning_rate": 0.0003901631420623437,
|
| 2265 |
+
"loss": 2.8551,
|
| 2266 |
+
"step": 319
|
| 2267 |
+
},
|
| 2268 |
+
{
|
| 2269 |
+
"epoch": 0.01836657283743587,
|
| 2270 |
+
"grad_norm": 0.20126426219940186,
|
| 2271 |
+
"learning_rate": 0.0003892824985316409,
|
| 2272 |
+
"loss": 2.7812,
|
| 2273 |
+
"step": 320
|
| 2274 |
+
},
|
| 2275 |
+
{
|
| 2276 |
+
"epoch": 0.018423968377552852,
|
| 2277 |
+
"grad_norm": 0.20343463122844696,
|
| 2278 |
+
"learning_rate": 0.0003884043508596093,
|
| 2279 |
+
"loss": 2.7959,
|
| 2280 |
+
"step": 321
|
| 2281 |
+
},
|
| 2282 |
+
{
|
| 2283 |
+
"epoch": 0.01848136391766984,
|
| 2284 |
+
"grad_norm": 0.22265484929084778,
|
| 2285 |
+
"learning_rate": 0.00038752873514346015,
|
| 2286 |
+
"loss": 2.8254,
|
| 2287 |
+
"step": 322
|
| 2288 |
+
},
|
| 2289 |
+
{
|
| 2290 |
+
"epoch": 0.018538759457786828,
|
| 2291 |
+
"grad_norm": 0.20545947551727295,
|
| 2292 |
+
"learning_rate": 0.000386655687376326,
|
| 2293 |
+
"loss": 2.8166,
|
| 2294 |
+
"step": 323
|
| 2295 |
+
},
|
| 2296 |
+
{
|
| 2297 |
+
"epoch": 0.018596154997903815,
|
| 2298 |
+
"grad_norm": 0.17015507817268372,
|
| 2299 |
+
"learning_rate": 0.00038578524344578115,
|
| 2300 |
+
"loss": 2.806,
|
| 2301 |
+
"step": 324
|
| 2302 |
+
},
|
| 2303 |
+
{
|
| 2304 |
+
"epoch": 0.018653550538020803,
|
| 2305 |
+
"grad_norm": 0.19378258287906647,
|
| 2306 |
+
"learning_rate": 0.00038491743913236624,
|
| 2307 |
+
"loss": 2.7979,
|
| 2308 |
+
"step": 325
|
| 2309 |
+
},
|
| 2310 |
+
{
|
| 2311 |
+
"epoch": 0.01871094607813779,
|
| 2312 |
+
"grad_norm": 0.2112617790699005,
|
| 2313 |
+
"learning_rate": 0.0003840523101081177,
|
| 2314 |
+
"loss": 2.8149,
|
| 2315 |
+
"step": 326
|
| 2316 |
+
},
|
| 2317 |
+
{
|
| 2318 |
+
"epoch": 0.018768341618254775,
|
| 2319 |
+
"grad_norm": 0.18846029043197632,
|
| 2320 |
+
"learning_rate": 0.0003831898919351011,
|
| 2321 |
+
"loss": 2.8334,
|
| 2322 |
+
"step": 327
|
| 2323 |
+
},
|
| 2324 |
+
{
|
| 2325 |
+
"epoch": 0.018825737158371762,
|
| 2326 |
+
"grad_norm": 0.20672033727169037,
|
| 2327 |
+
"learning_rate": 0.00038233022006394976,
|
| 2328 |
+
"loss": 2.8061,
|
| 2329 |
+
"step": 328
|
| 2330 |
+
},
|
| 2331 |
+
{
|
| 2332 |
+
"epoch": 0.01888313269848875,
|
| 2333 |
+
"grad_norm": 0.2700256109237671,
|
| 2334 |
+
"learning_rate": 0.00038147332983240717,
|
| 2335 |
+
"loss": 2.8101,
|
| 2336 |
+
"step": 329
|
| 2337 |
+
},
|
| 2338 |
+
{
|
| 2339 |
+
"epoch": 0.018940528238605737,
|
| 2340 |
+
"grad_norm": 0.16990099847316742,
|
| 2341 |
+
"learning_rate": 0.00038061925646387467,
|
| 2342 |
+
"loss": 2.8227,
|
| 2343 |
+
"step": 330
|
| 2344 |
+
},
|
| 2345 |
+
{
|
| 2346 |
+
"epoch": 0.018997923778722725,
|
| 2347 |
+
"grad_norm": 0.2140357792377472,
|
| 2348 |
+
"learning_rate": 0.0003797680350659631,
|
| 2349 |
+
"loss": 2.8018,
|
| 2350 |
+
"step": 331
|
| 2351 |
+
},
|
| 2352 |
+
{
|
| 2353 |
+
"epoch": 0.019055319318839713,
|
| 2354 |
+
"grad_norm": 0.2538260221481323,
|
| 2355 |
+
"learning_rate": 0.0003789197006290502,
|
| 2356 |
+
"loss": 2.7725,
|
| 2357 |
+
"step": 332
|
| 2358 |
+
},
|
| 2359 |
+
{
|
| 2360 |
+
"epoch": 0.0191127148589567,
|
| 2361 |
+
"grad_norm": 0.1694011092185974,
|
| 2362 |
+
"learning_rate": 0.0003780742880248419,
|
| 2363 |
+
"loss": 2.7973,
|
| 2364 |
+
"step": 333
|
| 2365 |
+
},
|
| 2366 |
+
{
|
| 2367 |
+
"epoch": 0.019170110399073684,
|
| 2368 |
+
"grad_norm": 0.2092764526605606,
|
| 2369 |
+
"learning_rate": 0.0003772318320049391,
|
| 2370 |
+
"loss": 2.8256,
|
| 2371 |
+
"step": 334
|
| 2372 |
+
},
|
| 2373 |
+
{
|
| 2374 |
+
"epoch": 0.019227505939190672,
|
| 2375 |
+
"grad_norm": 0.22675682604312897,
|
| 2376 |
+
"learning_rate": 0.0003763923671994093,
|
| 2377 |
+
"loss": 2.8092,
|
| 2378 |
+
"step": 335
|
| 2379 |
+
},
|
| 2380 |
+
{
|
| 2381 |
+
"epoch": 0.01928490147930766,
|
| 2382 |
+
"grad_norm": 0.20571155846118927,
|
| 2383 |
+
"learning_rate": 0.0003755559281153625,
|
| 2384 |
+
"loss": 2.8176,
|
| 2385 |
+
"step": 336
|
| 2386 |
+
},
|
| 2387 |
+
{
|
| 2388 |
+
"epoch": 0.019342297019424647,
|
| 2389 |
+
"grad_norm": 0.18606650829315186,
|
| 2390 |
+
"learning_rate": 0.0003747225491355334,
|
| 2391 |
+
"loss": 2.8019,
|
| 2392 |
+
"step": 337
|
| 2393 |
+
},
|
| 2394 |
+
{
|
| 2395 |
+
"epoch": 0.019399692559541635,
|
| 2396 |
+
"grad_norm": 0.19859890639781952,
|
| 2397 |
+
"learning_rate": 0.00037389226451686763,
|
| 2398 |
+
"loss": 2.8036,
|
| 2399 |
+
"step": 338
|
| 2400 |
+
},
|
| 2401 |
+
{
|
| 2402 |
+
"epoch": 0.019457088099658622,
|
| 2403 |
+
"grad_norm": 0.1632896512746811,
|
| 2404 |
+
"learning_rate": 0.00037306510838911404,
|
| 2405 |
+
"loss": 2.797,
|
| 2406 |
+
"step": 339
|
| 2407 |
+
},
|
| 2408 |
+
{
|
| 2409 |
+
"epoch": 0.01951448363977561,
|
| 2410 |
+
"grad_norm": 0.17494754493236542,
|
| 2411 |
+
"learning_rate": 0.00037224111475342116,
|
| 2412 |
+
"loss": 2.8152,
|
| 2413 |
+
"step": 340
|
| 2414 |
+
},
|
| 2415 |
+
{
|
| 2416 |
+
"epoch": 0.019571879179892594,
|
| 2417 |
+
"grad_norm": 0.20659732818603516,
|
| 2418 |
+
"learning_rate": 0.00037142031748094016,
|
| 2419 |
+
"loss": 2.8061,
|
| 2420 |
+
"step": 341
|
| 2421 |
+
},
|
| 2422 |
+
{
|
| 2423 |
+
"epoch": 0.019629274720009582,
|
| 2424 |
+
"grad_norm": 0.18716713786125183,
|
| 2425 |
+
"learning_rate": 0.00037060275031143184,
|
| 2426 |
+
"loss": 2.8419,
|
| 2427 |
+
"step": 342
|
| 2428 |
+
},
|
| 2429 |
+
{
|
| 2430 |
+
"epoch": 0.01968667026012657,
|
| 2431 |
+
"grad_norm": 0.2575749158859253,
|
| 2432 |
+
"learning_rate": 0.0003697884468518805,
|
| 2433 |
+
"loss": 2.7814,
|
| 2434 |
+
"step": 343
|
| 2435 |
+
},
|
| 2436 |
+
{
|
| 2437 |
+
"epoch": 0.019744065800243557,
|
| 2438 |
+
"grad_norm": 0.19076134264469147,
|
| 2439 |
+
"learning_rate": 0.0003689774405751119,
|
| 2440 |
+
"loss": 2.797,
|
| 2441 |
+
"step": 344
|
| 2442 |
+
},
|
| 2443 |
+
{
|
| 2444 |
+
"epoch": 0.019801461340360545,
|
| 2445 |
+
"grad_norm": 0.19563442468643188,
|
| 2446 |
+
"learning_rate": 0.00036816976481841764,
|
| 2447 |
+
"loss": 2.8269,
|
| 2448 |
+
"step": 345
|
| 2449 |
+
},
|
| 2450 |
+
{
|
| 2451 |
+
"epoch": 0.019858856880477532,
|
| 2452 |
+
"grad_norm": 0.1790810525417328,
|
| 2453 |
+
"learning_rate": 0.0003673654527821846,
|
| 2454 |
+
"loss": 2.7856,
|
| 2455 |
+
"step": 346
|
| 2456 |
+
},
|
| 2457 |
+
{
|
| 2458 |
+
"epoch": 0.019916252420594516,
|
| 2459 |
+
"grad_norm": 0.2125868797302246,
|
| 2460 |
+
"learning_rate": 0.00036656453752853025,
|
| 2461 |
+
"loss": 2.7973,
|
| 2462 |
+
"step": 347
|
| 2463 |
+
},
|
| 2464 |
+
{
|
| 2465 |
+
"epoch": 0.019973647960711504,
|
| 2466 |
+
"grad_norm": 0.1454995572566986,
|
| 2467 |
+
"learning_rate": 0.00036576705197994376,
|
| 2468 |
+
"loss": 2.7869,
|
| 2469 |
+
"step": 348
|
| 2470 |
+
},
|
| 2471 |
+
{
|
| 2472 |
+
"epoch": 0.02003104350082849,
|
| 2473 |
+
"grad_norm": 0.2808379530906677,
|
| 2474 |
+
"learning_rate": 0.00036497302891793255,
|
| 2475 |
+
"loss": 2.7923,
|
| 2476 |
+
"step": 349
|
| 2477 |
+
},
|
| 2478 |
+
{
|
| 2479 |
+
"epoch": 0.02008843904094548,
|
| 2480 |
+
"grad_norm": 0.1776140034198761,
|
| 2481 |
+
"learning_rate": 0.0003641825009816745,
|
| 2482 |
+
"loss": 2.8194,
|
| 2483 |
+
"step": 350
|
| 2484 |
+
},
|
| 2485 |
+
{
|
| 2486 |
+
"epoch": 0.020145834581062467,
|
| 2487 |
+
"grad_norm": 0.22207793593406677,
|
| 2488 |
+
"learning_rate": 0.0003633955006666771,
|
| 2489 |
+
"loss": 2.8234,
|
| 2490 |
+
"step": 351
|
| 2491 |
+
},
|
| 2492 |
+
{
|
| 2493 |
+
"epoch": 0.020203230121179454,
|
| 2494 |
+
"grad_norm": 0.24642404913902283,
|
| 2495 |
+
"learning_rate": 0.0003626120603234406,
|
| 2496 |
+
"loss": 2.8351,
|
| 2497 |
+
"step": 352
|
| 2498 |
+
},
|
| 2499 |
+
{
|
| 2500 |
+
"epoch": 0.020260625661296442,
|
| 2501 |
+
"grad_norm": 0.24731726944446564,
|
| 2502 |
+
"learning_rate": 0.000361832212156129,
|
| 2503 |
+
"loss": 2.7983,
|
| 2504 |
+
"step": 353
|
| 2505 |
+
},
|
| 2506 |
+
{
|
| 2507 |
+
"epoch": 0.020318021201413426,
|
| 2508 |
+
"grad_norm": 0.21677981317043304,
|
| 2509 |
+
"learning_rate": 0.0003610559882212461,
|
| 2510 |
+
"loss": 2.8372,
|
| 2511 |
+
"step": 354
|
| 2512 |
+
},
|
| 2513 |
+
{
|
| 2514 |
+
"epoch": 0.020375416741530414,
|
| 2515 |
+
"grad_norm": 0.28350090980529785,
|
| 2516 |
+
"learning_rate": 0.00036028342042631755,
|
| 2517 |
+
"loss": 2.8138,
|
| 2518 |
+
"step": 355
|
| 2519 |
+
},
|
| 2520 |
+
{
|
| 2521 |
+
"epoch": 0.0204328122816474,
|
| 2522 |
+
"grad_norm": 0.22418756783008575,
|
| 2523 |
+
"learning_rate": 0.00035951454052857954,
|
| 2524 |
+
"loss": 2.7897,
|
| 2525 |
+
"step": 356
|
| 2526 |
+
},
|
| 2527 |
+
{
|
| 2528 |
+
"epoch": 0.02049020782176439,
|
| 2529 |
+
"grad_norm": 0.27765804529190063,
|
| 2530 |
+
"learning_rate": 0.000358749380133673,
|
| 2531 |
+
"loss": 2.8139,
|
| 2532 |
+
"step": 357
|
| 2533 |
+
},
|
| 2534 |
+
{
|
| 2535 |
+
"epoch": 0.020547603361881377,
|
| 2536 |
+
"grad_norm": 0.2694258391857147,
|
| 2537 |
+
"learning_rate": 0.000357987970694345,
|
| 2538 |
+
"loss": 2.7881,
|
| 2539 |
+
"step": 358
|
| 2540 |
+
},
|
| 2541 |
+
{
|
| 2542 |
+
"epoch": 0.020604998901998364,
|
| 2543 |
+
"grad_norm": 0.3746117055416107,
|
| 2544 |
+
"learning_rate": 0.00035723034350915525,
|
| 2545 |
+
"loss": 2.8108,
|
| 2546 |
+
"step": 359
|
| 2547 |
+
},
|
| 2548 |
+
{
|
| 2549 |
+
"epoch": 0.02066239444211535,
|
| 2550 |
+
"grad_norm": 0.22864773869514465,
|
| 2551 |
+
"learning_rate": 0.00035647652972119,
|
| 2552 |
+
"loss": 2.8102,
|
| 2553 |
+
"step": 360
|
| 2554 |
+
},
|
| 2555 |
+
{
|
| 2556 |
+
"epoch": 0.020719789982232336,
|
| 2557 |
+
"grad_norm": 0.2728801369667053,
|
| 2558 |
+
"learning_rate": 0.0003557265603167814,
|
| 2559 |
+
"loss": 2.8046,
|
| 2560 |
+
"step": 361
|
| 2561 |
+
},
|
| 2562 |
+
{
|
| 2563 |
+
"epoch": 0.020777185522349324,
|
| 2564 |
+
"grad_norm": 0.2561710774898529,
|
| 2565 |
+
"learning_rate": 0.0003549804661242345,
|
| 2566 |
+
"loss": 2.8242,
|
| 2567 |
+
"step": 362
|
| 2568 |
+
},
|
| 2569 |
+
{
|
| 2570 |
+
"epoch": 0.02083458106246631,
|
| 2571 |
+
"grad_norm": 0.26235631108283997,
|
| 2572 |
+
"learning_rate": 0.00035423827781255914,
|
| 2573 |
+
"loss": 2.847,
|
| 2574 |
+
"step": 363
|
| 2575 |
+
},
|
| 2576 |
+
{
|
| 2577 |
+
"epoch": 0.0208919766025833,
|
| 2578 |
+
"grad_norm": 0.24725806713104248,
|
| 2579 |
+
"learning_rate": 0.0003535000258902099,
|
| 2580 |
+
"loss": 2.7873,
|
| 2581 |
+
"step": 364
|
| 2582 |
+
},
|
| 2583 |
+
{
|
| 2584 |
+
"epoch": 0.020949372142700286,
|
| 2585 |
+
"grad_norm": 0.2562279999256134,
|
| 2586 |
+
"learning_rate": 0.0003527657407038317,
|
| 2587 |
+
"loss": 2.799,
|
| 2588 |
+
"step": 365
|
| 2589 |
+
},
|
| 2590 |
+
{
|
| 2591 |
+
"epoch": 0.021006767682817274,
|
| 2592 |
+
"grad_norm": 0.20368199050426483,
|
| 2593 |
+
"learning_rate": 0.00035203545243701266,
|
| 2594 |
+
"loss": 2.8011,
|
| 2595 |
+
"step": 366
|
| 2596 |
+
},
|
| 2597 |
+
{
|
| 2598 |
+
"epoch": 0.021064163222934258,
|
| 2599 |
+
"grad_norm": 0.25594958662986755,
|
| 2600 |
+
"learning_rate": 0.0003513091911090431,
|
| 2601 |
+
"loss": 2.8099,
|
| 2602 |
+
"step": 367
|
| 2603 |
+
},
|
| 2604 |
+
{
|
| 2605 |
+
"epoch": 0.021121558763051246,
|
| 2606 |
+
"grad_norm": 0.20084761083126068,
|
| 2607 |
+
"learning_rate": 0.00035058698657368154,
|
| 2608 |
+
"loss": 2.8249,
|
| 2609 |
+
"step": 368
|
| 2610 |
+
},
|
| 2611 |
+
{
|
| 2612 |
+
"epoch": 0.021178954303168233,
|
| 2613 |
+
"grad_norm": 0.24110020697116852,
|
| 2614 |
+
"learning_rate": 0.00034986886851792775,
|
| 2615 |
+
"loss": 2.8058,
|
| 2616 |
+
"step": 369
|
| 2617 |
+
},
|
| 2618 |
+
{
|
| 2619 |
+
"epoch": 0.02123634984328522,
|
| 2620 |
+
"grad_norm": 0.2016633003950119,
|
| 2621 |
+
"learning_rate": 0.0003491548664608024,
|
| 2622 |
+
"loss": 2.7935,
|
| 2623 |
+
"step": 370
|
| 2624 |
+
},
|
| 2625 |
+
{
|
| 2626 |
+
"epoch": 0.02129374538340221,
|
| 2627 |
+
"grad_norm": 0.2722468376159668,
|
| 2628 |
+
"learning_rate": 0.0003484450097521336,
|
| 2629 |
+
"loss": 2.8146,
|
| 2630 |
+
"step": 371
|
| 2631 |
+
},
|
| 2632 |
+
{
|
| 2633 |
+
"epoch": 0.021351140923519196,
|
| 2634 |
+
"grad_norm": 0.2089434564113617,
|
| 2635 |
+
"learning_rate": 0.0003477393275713501,
|
| 2636 |
+
"loss": 2.8231,
|
| 2637 |
+
"step": 372
|
| 2638 |
+
},
|
| 2639 |
+
{
|
| 2640 |
+
"epoch": 0.021408536463636184,
|
| 2641 |
+
"grad_norm": 0.24770453572273254,
|
| 2642 |
+
"learning_rate": 0.0003470378489262824,
|
| 2643 |
+
"loss": 2.7994,
|
| 2644 |
+
"step": 373
|
| 2645 |
+
},
|
| 2646 |
+
{
|
| 2647 |
+
"epoch": 0.021465932003753168,
|
| 2648 |
+
"grad_norm": 0.21104897558689117,
|
| 2649 |
+
"learning_rate": 0.00034634060265197026,
|
| 2650 |
+
"loss": 2.8189,
|
| 2651 |
+
"step": 374
|
| 2652 |
+
},
|
| 2653 |
+
{
|
| 2654 |
+
"epoch": 0.021523327543870156,
|
| 2655 |
+
"grad_norm": 0.23374824225902557,
|
| 2656 |
+
"learning_rate": 0.000345647617409477,
|
| 2657 |
+
"loss": 2.783,
|
| 2658 |
+
"step": 375
|
| 2659 |
+
},
|
| 2660 |
+
{
|
| 2661 |
+
"epoch": 0.021580723083987143,
|
| 2662 |
+
"grad_norm": 0.24334168434143066,
|
| 2663 |
+
"learning_rate": 0.00034495892168471176,
|
| 2664 |
+
"loss": 2.8092,
|
| 2665 |
+
"step": 376
|
| 2666 |
+
},
|
| 2667 |
+
{
|
| 2668 |
+
"epoch": 0.02163811862410413,
|
| 2669 |
+
"grad_norm": 0.22772932052612305,
|
| 2670 |
+
"learning_rate": 0.00034427454378725827,
|
| 2671 |
+
"loss": 2.8178,
|
| 2672 |
+
"step": 377
|
| 2673 |
+
},
|
| 2674 |
+
{
|
| 2675 |
+
"epoch": 0.02169551416422112,
|
| 2676 |
+
"grad_norm": 0.22545067965984344,
|
| 2677 |
+
"learning_rate": 0.00034359451184921125,
|
| 2678 |
+
"loss": 2.7961,
|
| 2679 |
+
"step": 378
|
| 2680 |
+
},
|
| 2681 |
+
{
|
| 2682 |
+
"epoch": 0.021752909704338106,
|
| 2683 |
+
"grad_norm": 0.2873929738998413,
|
| 2684 |
+
"learning_rate": 0.00034291885382402044,
|
| 2685 |
+
"loss": 2.8408,
|
| 2686 |
+
"step": 379
|
| 2687 |
+
},
|
| 2688 |
+
{
|
| 2689 |
+
"epoch": 0.02181030524445509,
|
| 2690 |
+
"grad_norm": 0.2099824994802475,
|
| 2691 |
+
"learning_rate": 0.00034224759748534083,
|
| 2692 |
+
"loss": 2.782,
|
| 2693 |
+
"step": 380
|
| 2694 |
+
},
|
| 2695 |
+
{
|
| 2696 |
+
"epoch": 0.021867700784572078,
|
| 2697 |
+
"grad_norm": 0.32221996784210205,
|
| 2698 |
+
"learning_rate": 0.0003415807704258913,
|
| 2699 |
+
"loss": 2.8337,
|
| 2700 |
+
"step": 381
|
| 2701 |
+
},
|
| 2702 |
+
{
|
| 2703 |
+
"epoch": 0.021925096324689065,
|
| 2704 |
+
"grad_norm": 0.2531490623950958,
|
| 2705 |
+
"learning_rate": 0.0003409184000563204,
|
| 2706 |
+
"loss": 2.8273,
|
| 2707 |
+
"step": 382
|
| 2708 |
+
},
|
| 2709 |
+
{
|
| 2710 |
+
"epoch": 0.021982491864806053,
|
| 2711 |
+
"grad_norm": 0.3075484037399292,
|
| 2712 |
+
"learning_rate": 0.00034026051360407973,
|
| 2713 |
+
"loss": 2.7805,
|
| 2714 |
+
"step": 383
|
| 2715 |
+
},
|
| 2716 |
+
{
|
| 2717 |
+
"epoch": 0.02203988740492304,
|
| 2718 |
+
"grad_norm": 0.2366313338279724,
|
| 2719 |
+
"learning_rate": 0.0003396071381123047,
|
| 2720 |
+
"loss": 2.8278,
|
| 2721 |
+
"step": 384
|
| 2722 |
+
},
|
| 2723 |
+
{
|
| 2724 |
+
"epoch": 0.022097282945040028,
|
| 2725 |
+
"grad_norm": 0.2348204106092453,
|
| 2726 |
+
"learning_rate": 0.00033895830043870266,
|
| 2727 |
+
"loss": 2.7922,
|
| 2728 |
+
"step": 385
|
| 2729 |
+
},
|
| 2730 |
+
{
|
| 2731 |
+
"epoch": 0.022154678485157016,
|
| 2732 |
+
"grad_norm": 0.28124627470970154,
|
| 2733 |
+
"learning_rate": 0.00033831402725444896,
|
| 2734 |
+
"loss": 2.8065,
|
| 2735 |
+
"step": 386
|
| 2736 |
+
},
|
| 2737 |
+
{
|
| 2738 |
+
"epoch": 0.022212074025274,
|
| 2739 |
+
"grad_norm": 0.1927008032798767,
|
| 2740 |
+
"learning_rate": 0.0003376743450430907,
|
| 2741 |
+
"loss": 2.7958,
|
| 2742 |
+
"step": 387
|
| 2743 |
+
},
|
| 2744 |
+
{
|
| 2745 |
+
"epoch": 0.022269469565390988,
|
| 2746 |
+
"grad_norm": 0.26325997710227966,
|
| 2747 |
+
"learning_rate": 0.0003370392800994583,
|
| 2748 |
+
"loss": 2.8313,
|
| 2749 |
+
"step": 388
|
| 2750 |
+
},
|
| 2751 |
+
{
|
| 2752 |
+
"epoch": 0.022326865105507975,
|
| 2753 |
+
"grad_norm": 0.23394963145256042,
|
| 2754 |
+
"learning_rate": 0.0003364088585285842,
|
| 2755 |
+
"loss": 2.8126,
|
| 2756 |
+
"step": 389
|
| 2757 |
+
},
|
| 2758 |
+
{
|
| 2759 |
+
"epoch": 0.022384260645624963,
|
| 2760 |
+
"grad_norm": 0.26055994629859924,
|
| 2761 |
+
"learning_rate": 0.00033578310624462983,
|
| 2762 |
+
"loss": 2.787,
|
| 2763 |
+
"step": 390
|
| 2764 |
+
},
|
| 2765 |
+
{
|
| 2766 |
+
"epoch": 0.02244165618574195,
|
| 2767 |
+
"grad_norm": 0.2207145392894745,
|
| 2768 |
+
"learning_rate": 0.0003351620489698208,
|
| 2769 |
+
"loss": 2.796,
|
| 2770 |
+
"step": 391
|
| 2771 |
+
},
|
| 2772 |
+
{
|
| 2773 |
+
"epoch": 0.022499051725858938,
|
| 2774 |
+
"grad_norm": 0.34231698513031006,
|
| 2775 |
+
"learning_rate": 0.0003345457122333891,
|
| 2776 |
+
"loss": 2.7951,
|
| 2777 |
+
"step": 392
|
| 2778 |
+
},
|
| 2779 |
+
{
|
| 2780 |
+
"epoch": 0.022556447265975922,
|
| 2781 |
+
"grad_norm": 0.22361671924591064,
|
| 2782 |
+
"learning_rate": 0.00033393412137052396,
|
| 2783 |
+
"loss": 2.8251,
|
| 2784 |
+
"step": 393
|
| 2785 |
+
},
|
| 2786 |
+
{
|
| 2787 |
+
"epoch": 0.02261384280609291,
|
| 2788 |
+
"grad_norm": 0.24573372304439545,
|
| 2789 |
+
"learning_rate": 0.0003333273015213304,
|
| 2790 |
+
"loss": 2.7899,
|
| 2791 |
+
"step": 394
|
| 2792 |
+
},
|
| 2793 |
+
{
|
| 2794 |
+
"epoch": 0.022671238346209897,
|
| 2795 |
+
"grad_norm": 0.22109688818454742,
|
| 2796 |
+
"learning_rate": 0.0003327252776297955,
|
| 2797 |
+
"loss": 2.8178,
|
| 2798 |
+
"step": 395
|
| 2799 |
+
},
|
| 2800 |
+
{
|
| 2801 |
+
"epoch": 0.022728633886326885,
|
| 2802 |
+
"grad_norm": 0.22289875149726868,
|
| 2803 |
+
"learning_rate": 0.00033212807444276364,
|
| 2804 |
+
"loss": 2.8053,
|
| 2805 |
+
"step": 396
|
| 2806 |
+
},
|
| 2807 |
+
{
|
| 2808 |
+
"epoch": 0.022786029426443873,
|
| 2809 |
+
"grad_norm": 0.21445147693157196,
|
| 2810 |
+
"learning_rate": 0.00033153571650891865,
|
| 2811 |
+
"loss": 2.7998,
|
| 2812 |
+
"step": 397
|
| 2813 |
+
},
|
| 2814 |
+
{
|
| 2815 |
+
"epoch": 0.02284342496656086,
|
| 2816 |
+
"grad_norm": 0.25061139464378357,
|
| 2817 |
+
"learning_rate": 0.00033094822817777514,
|
| 2818 |
+
"loss": 2.8055,
|
| 2819 |
+
"step": 398
|
| 2820 |
+
},
|
| 2821 |
+
{
|
| 2822 |
+
"epoch": 0.022900820506677848,
|
| 2823 |
+
"grad_norm": 0.24680854380130768,
|
| 2824 |
+
"learning_rate": 0.0003303656335986773,
|
| 2825 |
+
"loss": 2.8143,
|
| 2826 |
+
"step": 399
|
| 2827 |
+
},
|
| 2828 |
+
{
|
| 2829 |
+
"epoch": 0.022958216046794832,
|
| 2830 |
+
"grad_norm": 0.16644932329654694,
|
| 2831 |
+
"learning_rate": 0.0003297879567198065,
|
| 2832 |
+
"loss": 2.8192,
|
| 2833 |
+
"step": 400
|
| 2834 |
+
},
|
| 2835 |
+
{
|
| 2836 |
+
"epoch": 0.022958216046794832,
|
| 2837 |
+
"eval_loss": 2.738191604614258,
|
| 2838 |
+
"eval_runtime": 85.3252,
|
| 2839 |
+
"eval_samples_per_second": 50.571,
|
| 2840 |
+
"eval_steps_per_second": 12.646,
|
| 2841 |
+
"step": 400
|
| 2842 |
+
},
|
| 2843 |
+
{
|
| 2844 |
+
"epoch": 0.02301561158691182,
|
| 2845 |
+
"grad_norm": 0.2816384434700012,
|
| 2846 |
+
"learning_rate": 0.00032921522128719657,
|
| 2847 |
+
"loss": 2.8209,
|
| 2848 |
+
"step": 401
|
| 2849 |
+
},
|
| 2850 |
+
{
|
| 2851 |
+
"epoch": 0.023073007127028807,
|
| 2852 |
+
"grad_norm": 0.20395685732364655,
|
| 2853 |
+
"learning_rate": 0.00032864745084375783,
|
| 2854 |
+
"loss": 2.8021,
|
| 2855 |
+
"step": 402
|
| 2856 |
+
},
|
| 2857 |
+
{
|
| 2858 |
+
"epoch": 0.023130402667145795,
|
| 2859 |
+
"grad_norm": 0.24216794967651367,
|
| 2860 |
+
"learning_rate": 0.00032808466872830957,
|
| 2861 |
+
"loss": 2.8447,
|
| 2862 |
+
"step": 403
|
| 2863 |
+
},
|
| 2864 |
+
{
|
| 2865 |
+
"epoch": 0.023187798207262782,
|
| 2866 |
+
"grad_norm": 0.2526738941669464,
|
| 2867 |
+
"learning_rate": 0.00032752689807462017,
|
| 2868 |
+
"loss": 2.7906,
|
| 2869 |
+
"step": 404
|
| 2870 |
+
},
|
| 2871 |
+
{
|
| 2872 |
+
"epoch": 0.02324519374737977,
|
| 2873 |
+
"grad_norm": 0.21725283563137054,
|
| 2874 |
+
"learning_rate": 0.0003269741618104566,
|
| 2875 |
+
"loss": 2.7943,
|
| 2876 |
+
"step": 405
|
| 2877 |
+
},
|
| 2878 |
+
{
|
| 2879 |
+
"epoch": 0.023302589287496758,
|
| 2880 |
+
"grad_norm": 0.2765718102455139,
|
| 2881 |
+
"learning_rate": 0.00032642648265664175,
|
| 2882 |
+
"loss": 2.8109,
|
| 2883 |
+
"step": 406
|
| 2884 |
+
},
|
| 2885 |
+
{
|
| 2886 |
+
"epoch": 0.02335998482761374,
|
| 2887 |
+
"grad_norm": 0.20015880465507507,
|
| 2888 |
+
"learning_rate": 0.00032588388312612053,
|
| 2889 |
+
"loss": 2.8239,
|
| 2890 |
+
"step": 407
|
| 2891 |
+
},
|
| 2892 |
+
{
|
| 2893 |
+
"epoch": 0.02341738036773073,
|
| 2894 |
+
"grad_norm": 0.26865240931510925,
|
| 2895 |
+
"learning_rate": 0.0003253463855230344,
|
| 2896 |
+
"loss": 2.8279,
|
| 2897 |
+
"step": 408
|
| 2898 |
+
},
|
| 2899 |
+
{
|
| 2900 |
+
"epoch": 0.023474775907847717,
|
| 2901 |
+
"grad_norm": 0.23522211611270905,
|
| 2902 |
+
"learning_rate": 0.0003248140119418046,
|
| 2903 |
+
"loss": 2.8123,
|
| 2904 |
+
"step": 409
|
| 2905 |
+
},
|
| 2906 |
+
{
|
| 2907 |
+
"epoch": 0.023532171447964705,
|
| 2908 |
+
"grad_norm": 0.2388644963502884,
|
| 2909 |
+
"learning_rate": 0.0003242867842662239,
|
| 2910 |
+
"loss": 2.8057,
|
| 2911 |
+
"step": 410
|
| 2912 |
+
},
|
| 2913 |
+
{
|
| 2914 |
+
"epoch": 0.023589566988081692,
|
| 2915 |
+
"grad_norm": 0.18323197960853577,
|
| 2916 |
+
"learning_rate": 0.00032376472416855703,
|
| 2917 |
+
"loss": 2.8193,
|
| 2918 |
+
"step": 411
|
| 2919 |
+
},
|
| 2920 |
+
{
|
| 2921 |
+
"epoch": 0.02364696252819868,
|
| 2922 |
+
"grad_norm": 0.24734856188297272,
|
| 2923 |
+
"learning_rate": 0.00032324785310864983,
|
| 2924 |
+
"loss": 2.7924,
|
| 2925 |
+
"step": 412
|
| 2926 |
+
},
|
| 2927 |
+
{
|
| 2928 |
+
"epoch": 0.023704358068315664,
|
| 2929 |
+
"grad_norm": 0.1722363829612732,
|
| 2930 |
+
"learning_rate": 0.0003227361923330471,
|
| 2931 |
+
"loss": 2.8242,
|
| 2932 |
+
"step": 413
|
| 2933 |
+
},
|
| 2934 |
+
{
|
| 2935 |
+
"epoch": 0.02376175360843265,
|
| 2936 |
+
"grad_norm": 0.2052358090877533,
|
| 2937 |
+
"learning_rate": 0.00032222976287411934,
|
| 2938 |
+
"loss": 2.8129,
|
| 2939 |
+
"step": 414
|
| 2940 |
+
},
|
| 2941 |
+
{
|
| 2942 |
+
"epoch": 0.02381914914854964,
|
| 2943 |
+
"grad_norm": 0.2536105811595917,
|
| 2944 |
+
"learning_rate": 0.00032172858554919807,
|
| 2945 |
+
"loss": 2.8207,
|
| 2946 |
+
"step": 415
|
| 2947 |
+
},
|
| 2948 |
+
{
|
| 2949 |
+
"epoch": 0.023876544688666627,
|
| 2950 |
+
"grad_norm": 0.23084022104740143,
|
| 2951 |
+
"learning_rate": 0.00032123268095972005,
|
| 2952 |
+
"loss": 2.8156,
|
| 2953 |
+
"step": 416
|
| 2954 |
+
},
|
| 2955 |
+
{
|
| 2956 |
+
"epoch": 0.023933940228783614,
|
| 2957 |
+
"grad_norm": 0.28741586208343506,
|
| 2958 |
+
"learning_rate": 0.00032074206949038073,
|
| 2959 |
+
"loss": 2.8008,
|
| 2960 |
+
"step": 417
|
| 2961 |
+
},
|
| 2962 |
+
{
|
| 2963 |
+
"epoch": 0.023991335768900602,
|
| 2964 |
+
"grad_norm": 0.2419297993183136,
|
| 2965 |
+
"learning_rate": 0.0003202567713082959,
|
| 2966 |
+
"loss": 2.8112,
|
| 2967 |
+
"step": 418
|
| 2968 |
+
},
|
| 2969 |
+
{
|
| 2970 |
+
"epoch": 0.02404873130901759,
|
| 2971 |
+
"grad_norm": 0.19744537770748138,
|
| 2972 |
+
"learning_rate": 0.0003197768063621732,
|
| 2973 |
+
"loss": 2.7894,
|
| 2974 |
+
"step": 419
|
| 2975 |
+
},
|
| 2976 |
+
{
|
| 2977 |
+
"epoch": 0.024106126849134574,
|
| 2978 |
+
"grad_norm": 0.22780993580818176,
|
| 2979 |
+
"learning_rate": 0.0003193021943814916,
|
| 2980 |
+
"loss": 2.8019,
|
| 2981 |
+
"step": 420
|
| 2982 |
+
},
|
| 2983 |
+
{
|
| 2984 |
+
"epoch": 0.02416352238925156,
|
| 2985 |
+
"grad_norm": 0.2176397144794464,
|
| 2986 |
+
"learning_rate": 0.00031883295487569063,
|
| 2987 |
+
"loss": 2.8183,
|
| 2988 |
+
"step": 421
|
| 2989 |
+
},
|
| 2990 |
+
{
|
| 2991 |
+
"epoch": 0.02422091792936855,
|
| 2992 |
+
"grad_norm": 0.23891203105449677,
|
| 2993 |
+
"learning_rate": 0.00031836910713336857,
|
| 2994 |
+
"loss": 2.8022,
|
| 2995 |
+
"step": 422
|
| 2996 |
+
},
|
| 2997 |
+
{
|
| 2998 |
+
"epoch": 0.024278313469485537,
|
| 2999 |
+
"grad_norm": 0.18507017195224762,
|
| 3000 |
+
"learning_rate": 0.0003179106702214893,
|
| 3001 |
+
"loss": 2.8013,
|
| 3002 |
+
"step": 423
|
| 3003 |
+
},
|
| 3004 |
+
{
|
| 3005 |
+
"epoch": 0.024335709009602524,
|
| 3006 |
+
"grad_norm": 0.20408926904201508,
|
| 3007 |
+
"learning_rate": 0.0003174576629845987,
|
| 3008 |
+
"loss": 2.8085,
|
| 3009 |
+
"step": 424
|
| 3010 |
+
},
|
| 3011 |
+
{
|
| 3012 |
+
"epoch": 0.024393104549719512,
|
| 3013 |
+
"grad_norm": 0.18055075407028198,
|
| 3014 |
+
"learning_rate": 0.00031701010404404996,
|
| 3015 |
+
"loss": 2.8341,
|
| 3016 |
+
"step": 425
|
| 3017 |
+
},
|
| 3018 |
+
{
|
| 3019 |
+
"epoch": 0.024450500089836496,
|
| 3020 |
+
"grad_norm": 0.22974956035614014,
|
| 3021 |
+
"learning_rate": 0.0003165680117972382,
|
| 3022 |
+
"loss": 2.8044,
|
| 3023 |
+
"step": 426
|
| 3024 |
+
},
|
| 3025 |
+
{
|
| 3026 |
+
"epoch": 0.024507895629953484,
|
| 3027 |
+
"grad_norm": 0.17688511312007904,
|
| 3028 |
+
"learning_rate": 0.00031613140441684413,
|
| 3029 |
+
"loss": 2.7866,
|
| 3030 |
+
"step": 427
|
| 3031 |
+
},
|
| 3032 |
+
{
|
| 3033 |
+
"epoch": 0.02456529117007047,
|
| 3034 |
+
"grad_norm": 0.22350828349590302,
|
| 3035 |
+
"learning_rate": 0.000315700299850087,
|
| 3036 |
+
"loss": 2.7939,
|
| 3037 |
+
"step": 428
|
| 3038 |
+
},
|
| 3039 |
+
{
|
| 3040 |
+
"epoch": 0.02462268671018746,
|
| 3041 |
+
"grad_norm": 0.2138863056898117,
|
| 3042 |
+
"learning_rate": 0.0003152747158179871,
|
| 3043 |
+
"loss": 2.8112,
|
| 3044 |
+
"step": 429
|
| 3045 |
+
},
|
| 3046 |
+
{
|
| 3047 |
+
"epoch": 0.024680082250304446,
|
| 3048 |
+
"grad_norm": 0.1666262447834015,
|
| 3049 |
+
"learning_rate": 0.0003148546698146371,
|
| 3050 |
+
"loss": 2.8464,
|
| 3051 |
+
"step": 430
|
| 3052 |
+
},
|
| 3053 |
+
{
|
| 3054 |
+
"epoch": 0.024737477790421434,
|
| 3055 |
+
"grad_norm": 0.23217864334583282,
|
| 3056 |
+
"learning_rate": 0.00031444017910648293,
|
| 3057 |
+
"loss": 2.8154,
|
| 3058 |
+
"step": 431
|
| 3059 |
+
},
|
| 3060 |
+
{
|
| 3061 |
+
"epoch": 0.02479487333053842,
|
| 3062 |
+
"grad_norm": 0.23967209458351135,
|
| 3063 |
+
"learning_rate": 0.00031403126073161424,
|
| 3064 |
+
"loss": 2.8068,
|
| 3065 |
+
"step": 432
|
| 3066 |
+
},
|
| 3067 |
+
{
|
| 3068 |
+
"epoch": 0.024852268870655406,
|
| 3069 |
+
"grad_norm": 0.2363416850566864,
|
| 3070 |
+
"learning_rate": 0.0003136279314990637,
|
| 3071 |
+
"loss": 2.832,
|
| 3072 |
+
"step": 433
|
| 3073 |
+
},
|
| 3074 |
+
{
|
| 3075 |
+
"epoch": 0.024909664410772393,
|
| 3076 |
+
"grad_norm": 0.20204566419124603,
|
| 3077 |
+
"learning_rate": 0.00031323020798811643,
|
| 3078 |
+
"loss": 2.8118,
|
| 3079 |
+
"step": 434
|
| 3080 |
+
},
|
| 3081 |
+
{
|
| 3082 |
+
"epoch": 0.02496705995088938,
|
| 3083 |
+
"grad_norm": 0.2645012438297272,
|
| 3084 |
+
"learning_rate": 0.00031283810654762816,
|
| 3085 |
+
"loss": 2.7988,
|
| 3086 |
+
"step": 435
|
| 3087 |
+
},
|
| 3088 |
+
{
|
| 3089 |
+
"epoch": 0.02502445549100637,
|
| 3090 |
+
"grad_norm": 0.31096434593200684,
|
| 3091 |
+
"learning_rate": 0.0003124516432953532,
|
| 3092 |
+
"loss": 2.8021,
|
| 3093 |
+
"step": 436
|
| 3094 |
+
},
|
| 3095 |
+
{
|
| 3096 |
+
"epoch": 0.025081851031123356,
|
| 3097 |
+
"grad_norm": 0.25740697979927063,
|
| 3098 |
+
"learning_rate": 0.00031207083411728236,
|
| 3099 |
+
"loss": 2.828,
|
| 3100 |
+
"step": 437
|
| 3101 |
+
},
|
| 3102 |
+
{
|
| 3103 |
+
"epoch": 0.025139246571240344,
|
| 3104 |
+
"grad_norm": 0.24895477294921875,
|
| 3105 |
+
"learning_rate": 0.00031169569466698937,
|
| 3106 |
+
"loss": 2.8073,
|
| 3107 |
+
"step": 438
|
| 3108 |
+
},
|
| 3109 |
+
{
|
| 3110 |
+
"epoch": 0.02519664211135733,
|
| 3111 |
+
"grad_norm": 0.2860502004623413,
|
| 3112 |
+
"learning_rate": 0.00031132624036498774,
|
| 3113 |
+
"loss": 2.8275,
|
| 3114 |
+
"step": 439
|
| 3115 |
+
},
|
| 3116 |
+
{
|
| 3117 |
+
"epoch": 0.025254037651474315,
|
| 3118 |
+
"grad_norm": 0.3134096562862396,
|
| 3119 |
+
"learning_rate": 0.00031096248639809674,
|
| 3120 |
+
"loss": 2.816,
|
| 3121 |
+
"step": 440
|
| 3122 |
+
},
|
| 3123 |
+
{
|
| 3124 |
+
"epoch": 0.025311433191591303,
|
| 3125 |
+
"grad_norm": 0.2185070812702179,
|
| 3126 |
+
"learning_rate": 0.0003106044477188172,
|
| 3127 |
+
"loss": 2.7799,
|
| 3128 |
+
"step": 441
|
| 3129 |
+
},
|
| 3130 |
+
{
|
| 3131 |
+
"epoch": 0.02536882873170829,
|
| 3132 |
+
"grad_norm": 0.3582714796066284,
|
| 3133 |
+
"learning_rate": 0.0003102521390447169,
|
| 3134 |
+
"loss": 2.7923,
|
| 3135 |
+
"step": 442
|
| 3136 |
+
},
|
| 3137 |
+
{
|
| 3138 |
+
"epoch": 0.02542622427182528,
|
| 3139 |
+
"grad_norm": 0.19494207203388214,
|
| 3140 |
+
"learning_rate": 0.00030990557485782553,
|
| 3141 |
+
"loss": 2.7999,
|
| 3142 |
+
"step": 443
|
| 3143 |
+
},
|
| 3144 |
+
{
|
| 3145 |
+
"epoch": 0.025483619811942266,
|
| 3146 |
+
"grad_norm": 0.2574940025806427,
|
| 3147 |
+
"learning_rate": 0.0003095647694040394,
|
| 3148 |
+
"loss": 2.8087,
|
| 3149 |
+
"step": 444
|
| 3150 |
+
},
|
| 3151 |
+
{
|
| 3152 |
+
"epoch": 0.025541015352059254,
|
| 3153 |
+
"grad_norm": 0.17501215636730194,
|
| 3154 |
+
"learning_rate": 0.0003092297366925359,
|
| 3155 |
+
"loss": 2.7817,
|
| 3156 |
+
"step": 445
|
| 3157 |
+
},
|
| 3158 |
+
{
|
| 3159 |
+
"epoch": 0.025598410892176238,
|
| 3160 |
+
"grad_norm": 0.4073377251625061,
|
| 3161 |
+
"learning_rate": 0.0003089004904951976,
|
| 3162 |
+
"loss": 2.813,
|
| 3163 |
+
"step": 446
|
| 3164 |
+
},
|
| 3165 |
+
{
|
| 3166 |
+
"epoch": 0.025655806432293225,
|
| 3167 |
+
"grad_norm": 0.21654489636421204,
|
| 3168 |
+
"learning_rate": 0.000308577044346046,
|
| 3169 |
+
"loss": 2.8165,
|
| 3170 |
+
"step": 447
|
| 3171 |
+
},
|
| 3172 |
+
{
|
| 3173 |
+
"epoch": 0.025713201972410213,
|
| 3174 |
+
"grad_norm": 0.26500189304351807,
|
| 3175 |
+
"learning_rate": 0.0003082594115406856,
|
| 3176 |
+
"loss": 2.8229,
|
| 3177 |
+
"step": 448
|
| 3178 |
+
},
|
| 3179 |
+
{
|
| 3180 |
+
"epoch": 0.0257705975125272,
|
| 3181 |
+
"grad_norm": 0.188262477517128,
|
| 3182 |
+
"learning_rate": 0.00030794760513575675,
|
| 3183 |
+
"loss": 2.8112,
|
| 3184 |
+
"step": 449
|
| 3185 |
+
},
|
| 3186 |
+
{
|
| 3187 |
+
"epoch": 0.025827993052644188,
|
| 3188 |
+
"grad_norm": 0.3432970643043518,
|
| 3189 |
+
"learning_rate": 0.00030764163794839966,
|
| 3190 |
+
"loss": 2.8241,
|
| 3191 |
+
"step": 450
|
| 3192 |
+
},
|
| 3193 |
+
{
|
| 3194 |
+
"epoch": 0.025885388592761176,
|
| 3195 |
+
"grad_norm": 0.23415225744247437,
|
| 3196 |
+
"learning_rate": 0.0003073415225557269,
|
| 3197 |
+
"loss": 2.8039,
|
| 3198 |
+
"step": 451
|
| 3199 |
+
},
|
| 3200 |
+
{
|
| 3201 |
+
"epoch": 0.025942784132878163,
|
| 3202 |
+
"grad_norm": 0.2670385241508484,
|
| 3203 |
+
"learning_rate": 0.0003070472712943069,
|
| 3204 |
+
"loss": 2.8215,
|
| 3205 |
+
"step": 452
|
| 3206 |
+
},
|
| 3207 |
+
{
|
| 3208 |
+
"epoch": 0.026000179672995147,
|
| 3209 |
+
"grad_norm": 0.17434735596179962,
|
| 3210 |
+
"learning_rate": 0.00030675889625965646,
|
| 3211 |
+
"loss": 2.8352,
|
| 3212 |
+
"step": 453
|
| 3213 |
+
},
|
| 3214 |
+
{
|
| 3215 |
+
"epoch": 0.026057575213112135,
|
| 3216 |
+
"grad_norm": 0.2789264917373657,
|
| 3217 |
+
"learning_rate": 0.0003064764093057437,
|
| 3218 |
+
"loss": 2.7856,
|
| 3219 |
+
"step": 454
|
| 3220 |
+
},
|
| 3221 |
+
{
|
| 3222 |
+
"epoch": 0.026114970753229123,
|
| 3223 |
+
"grad_norm": 0.2666022479534149,
|
| 3224 |
+
"learning_rate": 0.0003061998220445009,
|
| 3225 |
+
"loss": 2.8063,
|
| 3226 |
+
"step": 455
|
| 3227 |
+
},
|
| 3228 |
+
{
|
| 3229 |
+
"epoch": 0.02617236629334611,
|
| 3230 |
+
"grad_norm": 0.22438260912895203,
|
| 3231 |
+
"learning_rate": 0.00030592914584534706,
|
| 3232 |
+
"loss": 2.7783,
|
| 3233 |
+
"step": 456
|
| 3234 |
+
},
|
| 3235 |
+
{
|
| 3236 |
+
"epoch": 0.026229761833463098,
|
| 3237 |
+
"grad_norm": 0.2177169770002365,
|
| 3238 |
+
"learning_rate": 0.00030566439183472063,
|
| 3239 |
+
"loss": 2.786,
|
| 3240 |
+
"step": 457
|
| 3241 |
+
},
|
| 3242 |
+
{
|
| 3243 |
+
"epoch": 0.026287157373580086,
|
| 3244 |
+
"grad_norm": 0.22771142423152924,
|
| 3245 |
+
"learning_rate": 0.000305405570895622,
|
| 3246 |
+
"loss": 2.7881,
|
| 3247 |
+
"step": 458
|
| 3248 |
+
},
|
| 3249 |
+
{
|
| 3250 |
+
"epoch": 0.02634455291369707,
|
| 3251 |
+
"grad_norm": 0.29228097200393677,
|
| 3252 |
+
"learning_rate": 0.00030515269366716613,
|
| 3253 |
+
"loss": 2.7876,
|
| 3254 |
+
"step": 459
|
| 3255 |
+
},
|
| 3256 |
+
{
|
| 3257 |
+
"epoch": 0.026401948453814057,
|
| 3258 |
+
"grad_norm": 0.18204721808433533,
|
| 3259 |
+
"learning_rate": 0.00030490577054414553,
|
| 3260 |
+
"loss": 2.8153,
|
| 3261 |
+
"step": 460
|
| 3262 |
+
},
|
| 3263 |
+
{
|
| 3264 |
+
"epoch": 0.026459343993931045,
|
| 3265 |
+
"grad_norm": 0.19830970466136932,
|
| 3266 |
+
"learning_rate": 0.0003046648116766027,
|
| 3267 |
+
"loss": 2.7884,
|
| 3268 |
+
"step": 461
|
| 3269 |
+
},
|
| 3270 |
+
{
|
| 3271 |
+
"epoch": 0.026516739534048032,
|
| 3272 |
+
"grad_norm": 0.17311398684978485,
|
| 3273 |
+
"learning_rate": 0.00030442982696941276,
|
| 3274 |
+
"loss": 2.8055,
|
| 3275 |
+
"step": 462
|
| 3276 |
+
},
|
| 3277 |
+
{
|
| 3278 |
+
"epoch": 0.02657413507416502,
|
| 3279 |
+
"grad_norm": 0.21194536983966827,
|
| 3280 |
+
"learning_rate": 0.0003042008260818768,
|
| 3281 |
+
"loss": 2.815,
|
| 3282 |
+
"step": 463
|
| 3283 |
+
},
|
| 3284 |
+
{
|
| 3285 |
+
"epoch": 0.026631530614282008,
|
| 3286 |
+
"grad_norm": 0.22366400063037872,
|
| 3287 |
+
"learning_rate": 0.0003039778184273243,
|
| 3288 |
+
"loss": 2.7994,
|
| 3289 |
+
"step": 464
|
| 3290 |
+
},
|
| 3291 |
+
{
|
| 3292 |
+
"epoch": 0.026688926154398995,
|
| 3293 |
+
"grad_norm": 0.17785237729549408,
|
| 3294 |
+
"learning_rate": 0.00030376081317272645,
|
| 3295 |
+
"loss": 2.8049,
|
| 3296 |
+
"step": 465
|
| 3297 |
+
},
|
| 3298 |
+
{
|
| 3299 |
+
"epoch": 0.02674632169451598,
|
| 3300 |
+
"grad_norm": 0.2285715490579605,
|
| 3301 |
+
"learning_rate": 0.00030354981923831934,
|
| 3302 |
+
"loss": 2.8105,
|
| 3303 |
+
"step": 466
|
| 3304 |
+
},
|
| 3305 |
+
{
|
| 3306 |
+
"epoch": 0.026803717234632967,
|
| 3307 |
+
"grad_norm": 0.17985928058624268,
|
| 3308 |
+
"learning_rate": 0.0003033448452972373,
|
| 3309 |
+
"loss": 2.8246,
|
| 3310 |
+
"step": 467
|
| 3311 |
+
},
|
| 3312 |
+
{
|
| 3313 |
+
"epoch": 0.026861112774749955,
|
| 3314 |
+
"grad_norm": 0.2026437669992447,
|
| 3315 |
+
"learning_rate": 0.000303145899775156,
|
| 3316 |
+
"loss": 2.8192,
|
| 3317 |
+
"step": 468
|
| 3318 |
+
},
|
| 3319 |
+
{
|
| 3320 |
+
"epoch": 0.026918508314866942,
|
| 3321 |
+
"grad_norm": 0.2605213522911072,
|
| 3322 |
+
"learning_rate": 0.0003029529908499469,
|
| 3323 |
+
"loss": 2.826,
|
| 3324 |
+
"step": 469
|
| 3325 |
+
},
|
| 3326 |
+
{
|
| 3327 |
+
"epoch": 0.02697590385498393,
|
| 3328 |
+
"grad_norm": 0.22592206299304962,
|
| 3329 |
+
"learning_rate": 0.00030276612645134017,
|
| 3330 |
+
"loss": 2.7987,
|
| 3331 |
+
"step": 470
|
| 3332 |
+
},
|
| 3333 |
+
{
|
| 3334 |
+
"epoch": 0.027033299395100917,
|
| 3335 |
+
"grad_norm": 0.2988434433937073,
|
| 3336 |
+
"learning_rate": 0.0003025853142605994,
|
| 3337 |
+
"loss": 2.826,
|
| 3338 |
+
"step": 471
|
| 3339 |
+
},
|
| 3340 |
+
{
|
| 3341 |
+
"epoch": 0.027090694935217905,
|
| 3342 |
+
"grad_norm": 0.2247052788734436,
|
| 3343 |
+
"learning_rate": 0.0003024105617102055,
|
| 3344 |
+
"loss": 2.815,
|
| 3345 |
+
"step": 472
|
| 3346 |
+
},
|
| 3347 |
+
{
|
| 3348 |
+
"epoch": 0.02714809047533489,
|
| 3349 |
+
"grad_norm": 0.26565778255462646,
|
| 3350 |
+
"learning_rate": 0.00030224187598355145,
|
| 3351 |
+
"loss": 2.8283,
|
| 3352 |
+
"step": 473
|
| 3353 |
+
},
|
| 3354 |
+
{
|
| 3355 |
+
"epoch": 0.027205486015451877,
|
| 3356 |
+
"grad_norm": 0.2834932804107666,
|
| 3357 |
+
"learning_rate": 0.00030207926401464675,
|
| 3358 |
+
"loss": 2.8088,
|
| 3359 |
+
"step": 474
|
| 3360 |
+
},
|
| 3361 |
+
{
|
| 3362 |
+
"epoch": 0.027262881555568864,
|
| 3363 |
+
"grad_norm": 0.2396688312292099,
|
| 3364 |
+
"learning_rate": 0.0003019227324878324,
|
| 3365 |
+
"loss": 2.8024,
|
| 3366 |
+
"step": 475
|
| 3367 |
+
},
|
| 3368 |
+
{
|
| 3369 |
+
"epoch": 0.027320277095685852,
|
| 3370 |
+
"grad_norm": 0.2600051760673523,
|
| 3371 |
+
"learning_rate": 0.0003017722878375066,
|
| 3372 |
+
"loss": 2.8258,
|
| 3373 |
+
"step": 476
|
| 3374 |
+
},
|
| 3375 |
+
{
|
| 3376 |
+
"epoch": 0.02737767263580284,
|
| 3377 |
+
"grad_norm": 0.26368406414985657,
|
| 3378 |
+
"learning_rate": 0.00030162793624785957,
|
| 3379 |
+
"loss": 2.7875,
|
| 3380 |
+
"step": 477
|
| 3381 |
+
},
|
| 3382 |
+
{
|
| 3383 |
+
"epoch": 0.027435068175919827,
|
| 3384 |
+
"grad_norm": 0.389852911233902,
|
| 3385 |
+
"learning_rate": 0.0003014896836526197,
|
| 3386 |
+
"loss": 2.8166,
|
| 3387 |
+
"step": 478
|
| 3388 |
+
},
|
| 3389 |
+
{
|
| 3390 |
+
"epoch": 0.02749246371603681,
|
| 3391 |
+
"grad_norm": 0.23984675109386444,
|
| 3392 |
+
"learning_rate": 0.0003013575357348098,
|
| 3393 |
+
"loss": 2.8025,
|
| 3394 |
+
"step": 479
|
| 3395 |
+
},
|
| 3396 |
+
{
|
| 3397 |
+
"epoch": 0.0275498592561538,
|
| 3398 |
+
"grad_norm": 0.24591901898384094,
|
| 3399 |
+
"learning_rate": 0.00030123149792651307,
|
| 3400 |
+
"loss": 2.7898,
|
| 3401 |
+
"step": 480
|
| 3402 |
+
},
|
| 3403 |
+
{
|
| 3404 |
+
"epoch": 0.027607254796270787,
|
| 3405 |
+
"grad_norm": 0.24797213077545166,
|
| 3406 |
+
"learning_rate": 0.00030111157540865026,
|
| 3407 |
+
"loss": 2.8291,
|
| 3408 |
+
"step": 481
|
| 3409 |
+
},
|
| 3410 |
+
{
|
| 3411 |
+
"epoch": 0.027664650336387774,
|
| 3412 |
+
"grad_norm": 0.2542579770088196,
|
| 3413 |
+
"learning_rate": 0.0003009977731107663,
|
| 3414 |
+
"loss": 2.7868,
|
| 3415 |
+
"step": 482
|
| 3416 |
+
},
|
| 3417 |
+
{
|
| 3418 |
+
"epoch": 0.027722045876504762,
|
| 3419 |
+
"grad_norm": 0.21780452132225037,
|
| 3420 |
+
"learning_rate": 0.00030089009571082794,
|
| 3421 |
+
"loss": 2.8051,
|
| 3422 |
+
"step": 483
|
| 3423 |
+
},
|
| 3424 |
+
{
|
| 3425 |
+
"epoch": 0.02777944141662175,
|
| 3426 |
+
"grad_norm": 0.2790198028087616,
|
| 3427 |
+
"learning_rate": 0.0003007885476350314,
|
| 3428 |
+
"loss": 2.8004,
|
| 3429 |
+
"step": 484
|
| 3430 |
+
},
|
| 3431 |
+
{
|
| 3432 |
+
"epoch": 0.027836836956738737,
|
| 3433 |
+
"grad_norm": 0.2793212831020355,
|
| 3434 |
+
"learning_rate": 0.00030069313305762025,
|
| 3435 |
+
"loss": 2.8077,
|
| 3436 |
+
"step": 485
|
| 3437 |
+
},
|
| 3438 |
+
{
|
| 3439 |
+
"epoch": 0.02789423249685572,
|
| 3440 |
+
"grad_norm": 0.2663847506046295,
|
| 3441 |
+
"learning_rate": 0.0003006038559007141,
|
| 3442 |
+
"loss": 2.805,
|
| 3443 |
+
"step": 486
|
| 3444 |
+
},
|
| 3445 |
+
{
|
| 3446 |
+
"epoch": 0.02795162803697271,
|
| 3447 |
+
"grad_norm": 0.2695571482181549,
|
| 3448 |
+
"learning_rate": 0.0003005207198341473,
|
| 3449 |
+
"loss": 2.8102,
|
| 3450 |
+
"step": 487
|
| 3451 |
+
},
|
| 3452 |
+
{
|
| 3453 |
+
"epoch": 0.028009023577089696,
|
| 3454 |
+
"grad_norm": 0.3027716875076294,
|
| 3455 |
+
"learning_rate": 0.0003004437282753177,
|
| 3456 |
+
"loss": 2.7944,
|
| 3457 |
+
"step": 488
|
| 3458 |
+
},
|
| 3459 |
+
{
|
| 3460 |
+
"epoch": 0.028066419117206684,
|
| 3461 |
+
"grad_norm": 0.25220444798469543,
|
| 3462 |
+
"learning_rate": 0.0003003728843890469,
|
| 3463 |
+
"loss": 2.781,
|
| 3464 |
+
"step": 489
|
| 3465 |
+
},
|
| 3466 |
+
{
|
| 3467 |
+
"epoch": 0.02812381465732367,
|
| 3468 |
+
"grad_norm": 0.2733742594718933,
|
| 3469 |
+
"learning_rate": 0.0003003081910874495,
|
| 3470 |
+
"loss": 2.8138,
|
| 3471 |
+
"step": 490
|
| 3472 |
+
},
|
| 3473 |
+
{
|
| 3474 |
+
"epoch": 0.02818121019744066,
|
| 3475 |
+
"grad_norm": 0.23873530328273773,
|
| 3476 |
+
"learning_rate": 0.00030024965102981387,
|
| 3477 |
+
"loss": 2.8017,
|
| 3478 |
+
"step": 491
|
| 3479 |
+
},
|
| 3480 |
+
{
|
| 3481 |
+
"epoch": 0.028238605737557643,
|
| 3482 |
+
"grad_norm": 0.29158100485801697,
|
| 3483 |
+
"learning_rate": 0.0003001972666224923,
|
| 3484 |
+
"loss": 2.8084,
|
| 3485 |
+
"step": 492
|
| 3486 |
+
},
|
| 3487 |
+
{
|
| 3488 |
+
"epoch": 0.02829600127767463,
|
| 3489 |
+
"grad_norm": 0.3079324960708618,
|
| 3490 |
+
"learning_rate": 0.00030015104001880274,
|
| 3491 |
+
"loss": 2.8061,
|
| 3492 |
+
"step": 493
|
| 3493 |
+
},
|
| 3494 |
+
{
|
| 3495 |
+
"epoch": 0.02835339681779162,
|
| 3496 |
+
"grad_norm": 0.2448122203350067,
|
| 3497 |
+
"learning_rate": 0.00030011097311893984,
|
| 3498 |
+
"loss": 2.7817,
|
| 3499 |
+
"step": 494
|
| 3500 |
+
},
|
| 3501 |
+
{
|
| 3502 |
+
"epoch": 0.028410792357908606,
|
| 3503 |
+
"grad_norm": 0.3495275378227234,
|
| 3504 |
+
"learning_rate": 0.00030007706756989683,
|
| 3505 |
+
"loss": 2.8053,
|
| 3506 |
+
"step": 495
|
| 3507 |
+
},
|
| 3508 |
+
{
|
| 3509 |
+
"epoch": 0.028468187898025594,
|
| 3510 |
+
"grad_norm": 0.19935691356658936,
|
| 3511 |
+
"learning_rate": 0.000300049324765398,
|
| 3512 |
+
"loss": 2.7985,
|
| 3513 |
+
"step": 496
|
| 3514 |
+
},
|
| 3515 |
+
{
|
| 3516 |
+
"epoch": 0.02852558343814258,
|
| 3517 |
+
"grad_norm": 0.30157798528671265,
|
| 3518 |
+
"learning_rate": 0.0003000277458458415,
|
| 3519 |
+
"loss": 2.8271,
|
| 3520 |
+
"step": 497
|
| 3521 |
+
},
|
| 3522 |
+
{
|
| 3523 |
+
"epoch": 0.02858297897825957,
|
| 3524 |
+
"grad_norm": 0.23343823850154877,
|
| 3525 |
+
"learning_rate": 0.00030001233169825214,
|
| 3526 |
+
"loss": 2.807,
|
| 3527 |
+
"step": 498
|
| 3528 |
+
},
|
| 3529 |
+
{
|
| 3530 |
+
"epoch": 0.028640374518376553,
|
| 3531 |
+
"grad_norm": 0.25404173135757446,
|
| 3532 |
+
"learning_rate": 0.0003000030829562451,
|
| 3533 |
+
"loss": 2.8072,
|
| 3534 |
+
"step": 499
|
| 3535 |
+
},
|
| 3536 |
+
{
|
| 3537 |
+
"epoch": 0.02869777005849354,
|
| 3538 |
+
"grad_norm": 0.28863540291786194,
|
| 3539 |
+
"learning_rate": 0.0003,
|
| 3540 |
+
"loss": 2.8088,
|
| 3541 |
+
"step": 500
|
| 3542 |
+
},
|
| 3543 |
+
{
|
| 3544 |
+
"epoch": 0.02869777005849354,
|
| 3545 |
+
"eval_loss": 2.735079288482666,
|
| 3546 |
+
"eval_runtime": 85.4355,
|
| 3547 |
+
"eval_samples_per_second": 50.506,
|
| 3548 |
+
"eval_steps_per_second": 12.629,
|
| 3549 |
+
"step": 500
|
| 3550 |
+
}
|
| 3551 |
+
],
|
| 3552 |
+
"logging_steps": 1,
|
| 3553 |
+
"max_steps": 500,
|
| 3554 |
+
"num_input_tokens_seen": 0,
|
| 3555 |
+
"num_train_epochs": 1,
|
| 3556 |
+
"save_steps": 150,
|
| 3557 |
+
"stateful_callbacks": {
|
| 3558 |
+
"TrainerControl": {
|
| 3559 |
+
"args": {
|
| 3560 |
+
"should_epoch_stop": false,
|
| 3561 |
+
"should_evaluate": false,
|
| 3562 |
+
"should_log": false,
|
| 3563 |
+
"should_save": true,
|
| 3564 |
+
"should_training_stop": true
|
| 3565 |
+
},
|
| 3566 |
+
"attributes": {}
|
| 3567 |
+
}
|
| 3568 |
+
},
|
| 3569 |
+
"total_flos": 4.69922551431168e+17,
|
| 3570 |
+
"train_batch_size": 22,
|
| 3571 |
+
"trial_name": null,
|
| 3572 |
+
"trial_params": null
|
| 3573 |
+
}
|
training_args.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:594db8e3ba17a8fa3661cdc5102444839e6fe80ed0f8414f52615396a149cc65
|
| 3 |
+
size 5368
|