Upload folder using huggingface_hub
Browse files- config.json +43 -0
- model.safetensors +3 -0
- optimizer.pt +3 -0
- rng_state.pth +3 -0
- scheduler.pt +3 -0
- trainer_state.json +1405 -0
- training_args.bin +3 -0
config.json
ADDED
|
@@ -0,0 +1,43 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"_name_or_path": "./smollm-135M",
|
| 3 |
+
"_num_labels": 3,
|
| 4 |
+
"architectures": [
|
| 5 |
+
"LlamaForSequenceClassification"
|
| 6 |
+
],
|
| 7 |
+
"attention_bias": false,
|
| 8 |
+
"attention_dropout": 0.0,
|
| 9 |
+
"bos_token_id": 0,
|
| 10 |
+
"eos_token_id": 0,
|
| 11 |
+
"head_dim": 64,
|
| 12 |
+
"hidden_act": "silu",
|
| 13 |
+
"hidden_size": 576,
|
| 14 |
+
"id2label": {
|
| 15 |
+
"0": "negative",
|
| 16 |
+
"1": "neutral",
|
| 17 |
+
"2": "positive"
|
| 18 |
+
},
|
| 19 |
+
"initializer_range": 0.02,
|
| 20 |
+
"intermediate_size": 1536,
|
| 21 |
+
"label2id": {
|
| 22 |
+
"negative": 0,
|
| 23 |
+
"neutral": 1,
|
| 24 |
+
"positive": 2
|
| 25 |
+
},
|
| 26 |
+
"max_position_embeddings": 2048,
|
| 27 |
+
"mlp_bias": false,
|
| 28 |
+
"model_type": "llama",
|
| 29 |
+
"num_attention_heads": 9,
|
| 30 |
+
"num_hidden_layers": 30,
|
| 31 |
+
"num_key_value_heads": 3,
|
| 32 |
+
"pad_token_id": 0,
|
| 33 |
+
"pretraining_tp": 1,
|
| 34 |
+
"problem_type": "single_label_classification",
|
| 35 |
+
"rms_norm_eps": 1e-05,
|
| 36 |
+
"rope_scaling": null,
|
| 37 |
+
"rope_theta": 10000.0,
|
| 38 |
+
"tie_word_embeddings": true,
|
| 39 |
+
"torch_dtype": "float32",
|
| 40 |
+
"transformers_version": "4.48.0",
|
| 41 |
+
"use_cache": true,
|
| 42 |
+
"vocab_size": 49152
|
| 43 |
+
}
|
model.safetensors
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:10bb69df21ac54718dec2196ab81e4475b41f8cedcab8728efdcf6d706466120
|
| 3 |
+
size 538097400
|
optimizer.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:5a5b1e201a3a7011b6d538c88e32709395bb13363b480ad94da54a894fab81d4
|
| 3 |
+
size 1076356555
|
rng_state.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:391d01d3aeb4a35151817d446e4ba0b9c8a04084ae1b1b66eda188a30729da0a
|
| 3 |
+
size 14455
|
scheduler.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:a41615828fcc7147e89faeb04aac00fba7451f6a1f160d02a50a764db727d7a7
|
| 3 |
+
size 1465
|
trainer_state.json
ADDED
|
@@ -0,0 +1,1405 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": 0.014622284099459648,
|
| 3 |
+
"best_model_checkpoint": "sentiment-distillation-smollm/checkpoint-1875",
|
| 4 |
+
"epoch": 3.0,
|
| 5 |
+
"eval_steps": 500,
|
| 6 |
+
"global_step": 1875,
|
| 7 |
+
"is_hyper_param_search": false,
|
| 8 |
+
"is_local_process_zero": true,
|
| 9 |
+
"is_world_process_zero": true,
|
| 10 |
+
"log_history": [
|
| 11 |
+
{
|
| 12 |
+
"epoch": 0.016,
|
| 13 |
+
"grad_norm": 9.476157188415527,
|
| 14 |
+
"learning_rate": 1.597444089456869e-06,
|
| 15 |
+
"loss": 1.074,
|
| 16 |
+
"step": 10
|
| 17 |
+
},
|
| 18 |
+
{
|
| 19 |
+
"epoch": 0.032,
|
| 20 |
+
"grad_norm": 10.569486618041992,
|
| 21 |
+
"learning_rate": 3.194888178913738e-06,
|
| 22 |
+
"loss": 1.0372,
|
| 23 |
+
"step": 20
|
| 24 |
+
},
|
| 25 |
+
{
|
| 26 |
+
"epoch": 0.048,
|
| 27 |
+
"grad_norm": 8.480586051940918,
|
| 28 |
+
"learning_rate": 4.792332268370607e-06,
|
| 29 |
+
"loss": 0.9365,
|
| 30 |
+
"step": 30
|
| 31 |
+
},
|
| 32 |
+
{
|
| 33 |
+
"epoch": 0.064,
|
| 34 |
+
"grad_norm": 6.626020908355713,
|
| 35 |
+
"learning_rate": 6.389776357827476e-06,
|
| 36 |
+
"loss": 0.8842,
|
| 37 |
+
"step": 40
|
| 38 |
+
},
|
| 39 |
+
{
|
| 40 |
+
"epoch": 0.08,
|
| 41 |
+
"grad_norm": 12.755849838256836,
|
| 42 |
+
"learning_rate": 7.987220447284345e-06,
|
| 43 |
+
"loss": 0.9098,
|
| 44 |
+
"step": 50
|
| 45 |
+
},
|
| 46 |
+
{
|
| 47 |
+
"epoch": 0.096,
|
| 48 |
+
"grad_norm": 7.676151275634766,
|
| 49 |
+
"learning_rate": 9.584664536741214e-06,
|
| 50 |
+
"loss": 0.7348,
|
| 51 |
+
"step": 60
|
| 52 |
+
},
|
| 53 |
+
{
|
| 54 |
+
"epoch": 0.112,
|
| 55 |
+
"grad_norm": 5.426048278808594,
|
| 56 |
+
"learning_rate": 1.1182108626198083e-05,
|
| 57 |
+
"loss": 0.4936,
|
| 58 |
+
"step": 70
|
| 59 |
+
},
|
| 60 |
+
{
|
| 61 |
+
"epoch": 0.128,
|
| 62 |
+
"grad_norm": 3.9603002071380615,
|
| 63 |
+
"learning_rate": 1.2779552715654951e-05,
|
| 64 |
+
"loss": 0.3438,
|
| 65 |
+
"step": 80
|
| 66 |
+
},
|
| 67 |
+
{
|
| 68 |
+
"epoch": 0.144,
|
| 69 |
+
"grad_norm": 3.396127700805664,
|
| 70 |
+
"learning_rate": 1.4376996805111822e-05,
|
| 71 |
+
"loss": 0.1482,
|
| 72 |
+
"step": 90
|
| 73 |
+
},
|
| 74 |
+
{
|
| 75 |
+
"epoch": 0.16,
|
| 76 |
+
"grad_norm": 4.765463829040527,
|
| 77 |
+
"learning_rate": 1.597444089456869e-05,
|
| 78 |
+
"loss": 0.1855,
|
| 79 |
+
"step": 100
|
| 80 |
+
},
|
| 81 |
+
{
|
| 82 |
+
"epoch": 0.176,
|
| 83 |
+
"grad_norm": 0.04306459426879883,
|
| 84 |
+
"learning_rate": 1.757188498402556e-05,
|
| 85 |
+
"loss": 0.1788,
|
| 86 |
+
"step": 110
|
| 87 |
+
},
|
| 88 |
+
{
|
| 89 |
+
"epoch": 0.192,
|
| 90 |
+
"grad_norm": 0.04322272166609764,
|
| 91 |
+
"learning_rate": 1.9169329073482428e-05,
|
| 92 |
+
"loss": 0.1233,
|
| 93 |
+
"step": 120
|
| 94 |
+
},
|
| 95 |
+
{
|
| 96 |
+
"epoch": 0.208,
|
| 97 |
+
"grad_norm": 0.053582437336444855,
|
| 98 |
+
"learning_rate": 2.07667731629393e-05,
|
| 99 |
+
"loss": 0.0908,
|
| 100 |
+
"step": 130
|
| 101 |
+
},
|
| 102 |
+
{
|
| 103 |
+
"epoch": 0.224,
|
| 104 |
+
"grad_norm": 8.254273414611816,
|
| 105 |
+
"learning_rate": 2.2364217252396165e-05,
|
| 106 |
+
"loss": 0.1362,
|
| 107 |
+
"step": 140
|
| 108 |
+
},
|
| 109 |
+
{
|
| 110 |
+
"epoch": 0.24,
|
| 111 |
+
"grad_norm": 0.054730091243982315,
|
| 112 |
+
"learning_rate": 2.3961661341853036e-05,
|
| 113 |
+
"loss": 0.072,
|
| 114 |
+
"step": 150
|
| 115 |
+
},
|
| 116 |
+
{
|
| 117 |
+
"epoch": 0.256,
|
| 118 |
+
"grad_norm": 0.0763891264796257,
|
| 119 |
+
"learning_rate": 2.5559105431309903e-05,
|
| 120 |
+
"loss": 0.1823,
|
| 121 |
+
"step": 160
|
| 122 |
+
},
|
| 123 |
+
{
|
| 124 |
+
"epoch": 0.272,
|
| 125 |
+
"grad_norm": 6.117145538330078,
|
| 126 |
+
"learning_rate": 2.7156549520766773e-05,
|
| 127 |
+
"loss": 0.0963,
|
| 128 |
+
"step": 170
|
| 129 |
+
},
|
| 130 |
+
{
|
| 131 |
+
"epoch": 0.288,
|
| 132 |
+
"grad_norm": 1.8248779773712158,
|
| 133 |
+
"learning_rate": 2.8753993610223644e-05,
|
| 134 |
+
"loss": 0.2741,
|
| 135 |
+
"step": 180
|
| 136 |
+
},
|
| 137 |
+
{
|
| 138 |
+
"epoch": 0.304,
|
| 139 |
+
"grad_norm": 0.014617039822041988,
|
| 140 |
+
"learning_rate": 3.0351437699680514e-05,
|
| 141 |
+
"loss": 0.0691,
|
| 142 |
+
"step": 190
|
| 143 |
+
},
|
| 144 |
+
{
|
| 145 |
+
"epoch": 0.32,
|
| 146 |
+
"grad_norm": 2.316850423812866,
|
| 147 |
+
"learning_rate": 3.194888178913738e-05,
|
| 148 |
+
"loss": 0.0186,
|
| 149 |
+
"step": 200
|
| 150 |
+
},
|
| 151 |
+
{
|
| 152 |
+
"epoch": 0.336,
|
| 153 |
+
"grad_norm": 0.010957750491797924,
|
| 154 |
+
"learning_rate": 3.354632587859425e-05,
|
| 155 |
+
"loss": 0.1676,
|
| 156 |
+
"step": 210
|
| 157 |
+
},
|
| 158 |
+
{
|
| 159 |
+
"epoch": 0.352,
|
| 160 |
+
"grad_norm": 0.02525000460445881,
|
| 161 |
+
"learning_rate": 3.514376996805112e-05,
|
| 162 |
+
"loss": 0.1689,
|
| 163 |
+
"step": 220
|
| 164 |
+
},
|
| 165 |
+
{
|
| 166 |
+
"epoch": 0.368,
|
| 167 |
+
"grad_norm": 0.010558371432125568,
|
| 168 |
+
"learning_rate": 3.6741214057507985e-05,
|
| 169 |
+
"loss": 0.1759,
|
| 170 |
+
"step": 230
|
| 171 |
+
},
|
| 172 |
+
{
|
| 173 |
+
"epoch": 0.384,
|
| 174 |
+
"grad_norm": 0.009851646609604359,
|
| 175 |
+
"learning_rate": 3.8338658146964856e-05,
|
| 176 |
+
"loss": 0.1253,
|
| 177 |
+
"step": 240
|
| 178 |
+
},
|
| 179 |
+
{
|
| 180 |
+
"epoch": 0.4,
|
| 181 |
+
"grad_norm": 0.009924142621457577,
|
| 182 |
+
"learning_rate": 3.9936102236421726e-05,
|
| 183 |
+
"loss": 0.1141,
|
| 184 |
+
"step": 250
|
| 185 |
+
},
|
| 186 |
+
{
|
| 187 |
+
"epoch": 0.416,
|
| 188 |
+
"grad_norm": 0.00981380045413971,
|
| 189 |
+
"learning_rate": 4.15335463258786e-05,
|
| 190 |
+
"loss": 0.1227,
|
| 191 |
+
"step": 260
|
| 192 |
+
},
|
| 193 |
+
{
|
| 194 |
+
"epoch": 0.432,
|
| 195 |
+
"grad_norm": 4.685519218444824,
|
| 196 |
+
"learning_rate": 4.313099041533547e-05,
|
| 197 |
+
"loss": 0.2,
|
| 198 |
+
"step": 270
|
| 199 |
+
},
|
| 200 |
+
{
|
| 201 |
+
"epoch": 0.448,
|
| 202 |
+
"grad_norm": 1.1163793802261353,
|
| 203 |
+
"learning_rate": 4.472843450479233e-05,
|
| 204 |
+
"loss": 0.073,
|
| 205 |
+
"step": 280
|
| 206 |
+
},
|
| 207 |
+
{
|
| 208 |
+
"epoch": 0.464,
|
| 209 |
+
"grad_norm": 0.01022783387452364,
|
| 210 |
+
"learning_rate": 4.632587859424921e-05,
|
| 211 |
+
"loss": 0.1287,
|
| 212 |
+
"step": 290
|
| 213 |
+
},
|
| 214 |
+
{
|
| 215 |
+
"epoch": 0.48,
|
| 216 |
+
"grad_norm": 0.018457185477018356,
|
| 217 |
+
"learning_rate": 4.792332268370607e-05,
|
| 218 |
+
"loss": 0.0794,
|
| 219 |
+
"step": 300
|
| 220 |
+
},
|
| 221 |
+
{
|
| 222 |
+
"epoch": 0.496,
|
| 223 |
+
"grad_norm": 10.666505813598633,
|
| 224 |
+
"learning_rate": 4.952076677316294e-05,
|
| 225 |
+
"loss": 0.17,
|
| 226 |
+
"step": 310
|
| 227 |
+
},
|
| 228 |
+
{
|
| 229 |
+
"epoch": 0.512,
|
| 230 |
+
"grad_norm": 2.132030725479126,
|
| 231 |
+
"learning_rate": 4.987553342816501e-05,
|
| 232 |
+
"loss": 0.0952,
|
| 233 |
+
"step": 320
|
| 234 |
+
},
|
| 235 |
+
{
|
| 236 |
+
"epoch": 0.528,
|
| 237 |
+
"grad_norm": 11.730767250061035,
|
| 238 |
+
"learning_rate": 4.969772403982931e-05,
|
| 239 |
+
"loss": 0.1995,
|
| 240 |
+
"step": 330
|
| 241 |
+
},
|
| 242 |
+
{
|
| 243 |
+
"epoch": 0.544,
|
| 244 |
+
"grad_norm": 0.01148867979645729,
|
| 245 |
+
"learning_rate": 4.95199146514936e-05,
|
| 246 |
+
"loss": 0.1019,
|
| 247 |
+
"step": 340
|
| 248 |
+
},
|
| 249 |
+
{
|
| 250 |
+
"epoch": 0.56,
|
| 251 |
+
"grad_norm": 0.047348447144031525,
|
| 252 |
+
"learning_rate": 4.9342105263157894e-05,
|
| 253 |
+
"loss": 0.0706,
|
| 254 |
+
"step": 350
|
| 255 |
+
},
|
| 256 |
+
{
|
| 257 |
+
"epoch": 0.576,
|
| 258 |
+
"grad_norm": 0.010151210241019726,
|
| 259 |
+
"learning_rate": 4.916429587482219e-05,
|
| 260 |
+
"loss": 0.0878,
|
| 261 |
+
"step": 360
|
| 262 |
+
},
|
| 263 |
+
{
|
| 264 |
+
"epoch": 0.592,
|
| 265 |
+
"grad_norm": 0.4587936997413635,
|
| 266 |
+
"learning_rate": 4.8986486486486486e-05,
|
| 267 |
+
"loss": 0.0864,
|
| 268 |
+
"step": 370
|
| 269 |
+
},
|
| 270 |
+
{
|
| 271 |
+
"epoch": 0.608,
|
| 272 |
+
"grad_norm": 0.1231539249420166,
|
| 273 |
+
"learning_rate": 4.8808677098150786e-05,
|
| 274 |
+
"loss": 0.0096,
|
| 275 |
+
"step": 380
|
| 276 |
+
},
|
| 277 |
+
{
|
| 278 |
+
"epoch": 0.624,
|
| 279 |
+
"grad_norm": 0.005615161266177893,
|
| 280 |
+
"learning_rate": 4.863086770981508e-05,
|
| 281 |
+
"loss": 0.1351,
|
| 282 |
+
"step": 390
|
| 283 |
+
},
|
| 284 |
+
{
|
| 285 |
+
"epoch": 0.64,
|
| 286 |
+
"grad_norm": 0.02301601506769657,
|
| 287 |
+
"learning_rate": 4.845305832147938e-05,
|
| 288 |
+
"loss": 0.0363,
|
| 289 |
+
"step": 400
|
| 290 |
+
},
|
| 291 |
+
{
|
| 292 |
+
"epoch": 0.656,
|
| 293 |
+
"grad_norm": 3.3105900287628174,
|
| 294 |
+
"learning_rate": 4.827524893314367e-05,
|
| 295 |
+
"loss": 0.0764,
|
| 296 |
+
"step": 410
|
| 297 |
+
},
|
| 298 |
+
{
|
| 299 |
+
"epoch": 0.672,
|
| 300 |
+
"grad_norm": 0.0835251584649086,
|
| 301 |
+
"learning_rate": 4.809743954480797e-05,
|
| 302 |
+
"loss": 0.0238,
|
| 303 |
+
"step": 420
|
| 304 |
+
},
|
| 305 |
+
{
|
| 306 |
+
"epoch": 0.688,
|
| 307 |
+
"grad_norm": 0.0839819684624672,
|
| 308 |
+
"learning_rate": 4.7919630156472264e-05,
|
| 309 |
+
"loss": 0.1182,
|
| 310 |
+
"step": 430
|
| 311 |
+
},
|
| 312 |
+
{
|
| 313 |
+
"epoch": 0.704,
|
| 314 |
+
"grad_norm": 0.0160073135048151,
|
| 315 |
+
"learning_rate": 4.774182076813656e-05,
|
| 316 |
+
"loss": 0.1647,
|
| 317 |
+
"step": 440
|
| 318 |
+
},
|
| 319 |
+
{
|
| 320 |
+
"epoch": 0.72,
|
| 321 |
+
"grad_norm": 0.09903815388679504,
|
| 322 |
+
"learning_rate": 4.756401137980086e-05,
|
| 323 |
+
"loss": 0.138,
|
| 324 |
+
"step": 450
|
| 325 |
+
},
|
| 326 |
+
{
|
| 327 |
+
"epoch": 0.736,
|
| 328 |
+
"grad_norm": 8.14660930633545,
|
| 329 |
+
"learning_rate": 4.738620199146515e-05,
|
| 330 |
+
"loss": 0.0731,
|
| 331 |
+
"step": 460
|
| 332 |
+
},
|
| 333 |
+
{
|
| 334 |
+
"epoch": 0.752,
|
| 335 |
+
"grad_norm": 0.07308264076709747,
|
| 336 |
+
"learning_rate": 4.720839260312945e-05,
|
| 337 |
+
"loss": 0.017,
|
| 338 |
+
"step": 470
|
| 339 |
+
},
|
| 340 |
+
{
|
| 341 |
+
"epoch": 0.768,
|
| 342 |
+
"grad_norm": 0.8220946788787842,
|
| 343 |
+
"learning_rate": 4.703058321479374e-05,
|
| 344 |
+
"loss": 0.0299,
|
| 345 |
+
"step": 480
|
| 346 |
+
},
|
| 347 |
+
{
|
| 348 |
+
"epoch": 0.784,
|
| 349 |
+
"grad_norm": 0.007219772785902023,
|
| 350 |
+
"learning_rate": 4.685277382645804e-05,
|
| 351 |
+
"loss": 0.0162,
|
| 352 |
+
"step": 490
|
| 353 |
+
},
|
| 354 |
+
{
|
| 355 |
+
"epoch": 0.8,
|
| 356 |
+
"grad_norm": 0.009510455653071404,
|
| 357 |
+
"learning_rate": 4.6674964438122335e-05,
|
| 358 |
+
"loss": 0.0011,
|
| 359 |
+
"step": 500
|
| 360 |
+
},
|
| 361 |
+
{
|
| 362 |
+
"epoch": 0.816,
|
| 363 |
+
"grad_norm": 9.018925666809082,
|
| 364 |
+
"learning_rate": 4.6497155049786634e-05,
|
| 365 |
+
"loss": 0.1081,
|
| 366 |
+
"step": 510
|
| 367 |
+
},
|
| 368 |
+
{
|
| 369 |
+
"epoch": 0.832,
|
| 370 |
+
"grad_norm": 0.008953461423516273,
|
| 371 |
+
"learning_rate": 4.631934566145093e-05,
|
| 372 |
+
"loss": 0.0494,
|
| 373 |
+
"step": 520
|
| 374 |
+
},
|
| 375 |
+
{
|
| 376 |
+
"epoch": 0.848,
|
| 377 |
+
"grad_norm": 10.596187591552734,
|
| 378 |
+
"learning_rate": 4.614153627311522e-05,
|
| 379 |
+
"loss": 0.0158,
|
| 380 |
+
"step": 530
|
| 381 |
+
},
|
| 382 |
+
{
|
| 383 |
+
"epoch": 0.864,
|
| 384 |
+
"grad_norm": 2.3671252727508545,
|
| 385 |
+
"learning_rate": 4.596372688477952e-05,
|
| 386 |
+
"loss": 0.0816,
|
| 387 |
+
"step": 540
|
| 388 |
+
},
|
| 389 |
+
{
|
| 390 |
+
"epoch": 0.88,
|
| 391 |
+
"grad_norm": 0.002445698482915759,
|
| 392 |
+
"learning_rate": 4.578591749644381e-05,
|
| 393 |
+
"loss": 0.0042,
|
| 394 |
+
"step": 550
|
| 395 |
+
},
|
| 396 |
+
{
|
| 397 |
+
"epoch": 0.896,
|
| 398 |
+
"grad_norm": 11.779576301574707,
|
| 399 |
+
"learning_rate": 4.560810810810811e-05,
|
| 400 |
+
"loss": 0.1067,
|
| 401 |
+
"step": 560
|
| 402 |
+
},
|
| 403 |
+
{
|
| 404 |
+
"epoch": 0.912,
|
| 405 |
+
"grad_norm": 7.589051246643066,
|
| 406 |
+
"learning_rate": 4.5430298719772405e-05,
|
| 407 |
+
"loss": 0.0853,
|
| 408 |
+
"step": 570
|
| 409 |
+
},
|
| 410 |
+
{
|
| 411 |
+
"epoch": 0.928,
|
| 412 |
+
"grad_norm": 0.11888572573661804,
|
| 413 |
+
"learning_rate": 4.5252489331436705e-05,
|
| 414 |
+
"loss": 0.0317,
|
| 415 |
+
"step": 580
|
| 416 |
+
},
|
| 417 |
+
{
|
| 418 |
+
"epoch": 0.944,
|
| 419 |
+
"grad_norm": 0.07746365666389465,
|
| 420 |
+
"learning_rate": 4.5074679943101e-05,
|
| 421 |
+
"loss": 0.2303,
|
| 422 |
+
"step": 590
|
| 423 |
+
},
|
| 424 |
+
{
|
| 425 |
+
"epoch": 0.96,
|
| 426 |
+
"grad_norm": 0.004527238663285971,
|
| 427 |
+
"learning_rate": 4.489687055476529e-05,
|
| 428 |
+
"loss": 0.0665,
|
| 429 |
+
"step": 600
|
| 430 |
+
},
|
| 431 |
+
{
|
| 432 |
+
"epoch": 0.976,
|
| 433 |
+
"grad_norm": 4.120000839233398,
|
| 434 |
+
"learning_rate": 4.471906116642959e-05,
|
| 435 |
+
"loss": 0.0546,
|
| 436 |
+
"step": 610
|
| 437 |
+
},
|
| 438 |
+
{
|
| 439 |
+
"epoch": 0.992,
|
| 440 |
+
"grad_norm": 0.0016751989023759961,
|
| 441 |
+
"learning_rate": 4.4541251778093884e-05,
|
| 442 |
+
"loss": 0.0051,
|
| 443 |
+
"step": 620
|
| 444 |
+
},
|
| 445 |
+
{
|
| 446 |
+
"epoch": 1.0,
|
| 447 |
+
"eval_accuracy": 0.9939879759519038,
|
| 448 |
+
"eval_f1_macro": 0.9861065168559803,
|
| 449 |
+
"eval_f1_micro": 0.9939879759519038,
|
| 450 |
+
"eval_f1_weighted": 0.9939373007372426,
|
| 451 |
+
"eval_loss": 0.015998151153326035,
|
| 452 |
+
"eval_precision_macro": 0.9959514170040485,
|
| 453 |
+
"eval_precision_micro": 0.9939879759519038,
|
| 454 |
+
"eval_precision_weighted": 0.994060996486901,
|
| 455 |
+
"eval_recall_macro": 0.9770065284178188,
|
| 456 |
+
"eval_recall_micro": 0.9939879759519038,
|
| 457 |
+
"eval_recall_weighted": 0.9939879759519038,
|
| 458 |
+
"eval_runtime": 10.7994,
|
| 459 |
+
"eval_samples_per_second": 46.206,
|
| 460 |
+
"eval_steps_per_second": 2.963,
|
| 461 |
+
"step": 625
|
| 462 |
+
},
|
| 463 |
+
{
|
| 464 |
+
"epoch": 1.008,
|
| 465 |
+
"grad_norm": 4.363436222076416,
|
| 466 |
+
"learning_rate": 4.436344238975818e-05,
|
| 467 |
+
"loss": 0.0119,
|
| 468 |
+
"step": 630
|
| 469 |
+
},
|
| 470 |
+
{
|
| 471 |
+
"epoch": 1.024,
|
| 472 |
+
"grad_norm": 0.002567918971180916,
|
| 473 |
+
"learning_rate": 4.4185633001422476e-05,
|
| 474 |
+
"loss": 0.0736,
|
| 475 |
+
"step": 640
|
| 476 |
+
},
|
| 477 |
+
{
|
| 478 |
+
"epoch": 1.04,
|
| 479 |
+
"grad_norm": 0.00732471002265811,
|
| 480 |
+
"learning_rate": 4.4007823613086776e-05,
|
| 481 |
+
"loss": 0.0001,
|
| 482 |
+
"step": 650
|
| 483 |
+
},
|
| 484 |
+
{
|
| 485 |
+
"epoch": 1.056,
|
| 486 |
+
"grad_norm": 0.002863505156710744,
|
| 487 |
+
"learning_rate": 4.383001422475107e-05,
|
| 488 |
+
"loss": 0.0608,
|
| 489 |
+
"step": 660
|
| 490 |
+
},
|
| 491 |
+
{
|
| 492 |
+
"epoch": 1.072,
|
| 493 |
+
"grad_norm": 0.008520668372511864,
|
| 494 |
+
"learning_rate": 4.365220483641537e-05,
|
| 495 |
+
"loss": 0.0125,
|
| 496 |
+
"step": 670
|
| 497 |
+
},
|
| 498 |
+
{
|
| 499 |
+
"epoch": 1.088,
|
| 500 |
+
"grad_norm": 0.001538406009785831,
|
| 501 |
+
"learning_rate": 4.347439544807966e-05,
|
| 502 |
+
"loss": 0.0653,
|
| 503 |
+
"step": 680
|
| 504 |
+
},
|
| 505 |
+
{
|
| 506 |
+
"epoch": 1.104,
|
| 507 |
+
"grad_norm": 0.001671103062108159,
|
| 508 |
+
"learning_rate": 4.3296586059743954e-05,
|
| 509 |
+
"loss": 0.081,
|
| 510 |
+
"step": 690
|
| 511 |
+
},
|
| 512 |
+
{
|
| 513 |
+
"epoch": 1.12,
|
| 514 |
+
"grad_norm": 0.20056261122226715,
|
| 515 |
+
"learning_rate": 4.3118776671408254e-05,
|
| 516 |
+
"loss": 0.0057,
|
| 517 |
+
"step": 700
|
| 518 |
+
},
|
| 519 |
+
{
|
| 520 |
+
"epoch": 1.1360000000000001,
|
| 521 |
+
"grad_norm": 0.008405996486544609,
|
| 522 |
+
"learning_rate": 4.294096728307255e-05,
|
| 523 |
+
"loss": 0.0091,
|
| 524 |
+
"step": 710
|
| 525 |
+
},
|
| 526 |
+
{
|
| 527 |
+
"epoch": 1.152,
|
| 528 |
+
"grad_norm": 0.19218170642852783,
|
| 529 |
+
"learning_rate": 4.2763157894736847e-05,
|
| 530 |
+
"loss": 0.0019,
|
| 531 |
+
"step": 720
|
| 532 |
+
},
|
| 533 |
+
{
|
| 534 |
+
"epoch": 1.168,
|
| 535 |
+
"grad_norm": 0.07524432241916656,
|
| 536 |
+
"learning_rate": 4.258534850640114e-05,
|
| 537 |
+
"loss": 0.0952,
|
| 538 |
+
"step": 730
|
| 539 |
+
},
|
| 540 |
+
{
|
| 541 |
+
"epoch": 1.184,
|
| 542 |
+
"grad_norm": 0.0015570322284474969,
|
| 543 |
+
"learning_rate": 4.240753911806544e-05,
|
| 544 |
+
"loss": 0.0435,
|
| 545 |
+
"step": 740
|
| 546 |
+
},
|
| 547 |
+
{
|
| 548 |
+
"epoch": 1.2,
|
| 549 |
+
"grad_norm": 0.0036333040334284306,
|
| 550 |
+
"learning_rate": 4.222972972972973e-05,
|
| 551 |
+
"loss": 0.0207,
|
| 552 |
+
"step": 750
|
| 553 |
+
},
|
| 554 |
+
{
|
| 555 |
+
"epoch": 1.216,
|
| 556 |
+
"grad_norm": 9.78964900970459,
|
| 557 |
+
"learning_rate": 4.205192034139403e-05,
|
| 558 |
+
"loss": 0.1004,
|
| 559 |
+
"step": 760
|
| 560 |
+
},
|
| 561 |
+
{
|
| 562 |
+
"epoch": 1.232,
|
| 563 |
+
"grad_norm": 0.0014888375299051404,
|
| 564 |
+
"learning_rate": 4.187411095305832e-05,
|
| 565 |
+
"loss": 0.0671,
|
| 566 |
+
"step": 770
|
| 567 |
+
},
|
| 568 |
+
{
|
| 569 |
+
"epoch": 1.248,
|
| 570 |
+
"grad_norm": 0.023692140355706215,
|
| 571 |
+
"learning_rate": 4.169630156472262e-05,
|
| 572 |
+
"loss": 0.0026,
|
| 573 |
+
"step": 780
|
| 574 |
+
},
|
| 575 |
+
{
|
| 576 |
+
"epoch": 1.264,
|
| 577 |
+
"grad_norm": 0.004593148361891508,
|
| 578 |
+
"learning_rate": 4.151849217638692e-05,
|
| 579 |
+
"loss": 0.0423,
|
| 580 |
+
"step": 790
|
| 581 |
+
},
|
| 582 |
+
{
|
| 583 |
+
"epoch": 1.28,
|
| 584 |
+
"grad_norm": 0.003923716489225626,
|
| 585 |
+
"learning_rate": 4.134068278805121e-05,
|
| 586 |
+
"loss": 0.0236,
|
| 587 |
+
"step": 800
|
| 588 |
+
},
|
| 589 |
+
{
|
| 590 |
+
"epoch": 1.296,
|
| 591 |
+
"grad_norm": 0.002044517546892166,
|
| 592 |
+
"learning_rate": 4.116287339971551e-05,
|
| 593 |
+
"loss": 0.005,
|
| 594 |
+
"step": 810
|
| 595 |
+
},
|
| 596 |
+
{
|
| 597 |
+
"epoch": 1.312,
|
| 598 |
+
"grad_norm": 0.022982032969594002,
|
| 599 |
+
"learning_rate": 4.09850640113798e-05,
|
| 600 |
+
"loss": 0.0697,
|
| 601 |
+
"step": 820
|
| 602 |
+
},
|
| 603 |
+
{
|
| 604 |
+
"epoch": 1.328,
|
| 605 |
+
"grad_norm": 0.23058412969112396,
|
| 606 |
+
"learning_rate": 4.08072546230441e-05,
|
| 607 |
+
"loss": 0.1222,
|
| 608 |
+
"step": 830
|
| 609 |
+
},
|
| 610 |
+
{
|
| 611 |
+
"epoch": 1.3439999999999999,
|
| 612 |
+
"grad_norm": 0.0017761716153472662,
|
| 613 |
+
"learning_rate": 4.0629445234708395e-05,
|
| 614 |
+
"loss": 0.0022,
|
| 615 |
+
"step": 840
|
| 616 |
+
},
|
| 617 |
+
{
|
| 618 |
+
"epoch": 1.3599999999999999,
|
| 619 |
+
"grad_norm": 0.12850730121135712,
|
| 620 |
+
"learning_rate": 4.0451635846372695e-05,
|
| 621 |
+
"loss": 0.0189,
|
| 622 |
+
"step": 850
|
| 623 |
+
},
|
| 624 |
+
{
|
| 625 |
+
"epoch": 1.376,
|
| 626 |
+
"grad_norm": 0.0014099746476858854,
|
| 627 |
+
"learning_rate": 4.027382645803698e-05,
|
| 628 |
+
"loss": 0.0001,
|
| 629 |
+
"step": 860
|
| 630 |
+
},
|
| 631 |
+
{
|
| 632 |
+
"epoch": 1.392,
|
| 633 |
+
"grad_norm": 10.44194507598877,
|
| 634 |
+
"learning_rate": 4.009601706970128e-05,
|
| 635 |
+
"loss": 0.1423,
|
| 636 |
+
"step": 870
|
| 637 |
+
},
|
| 638 |
+
{
|
| 639 |
+
"epoch": 1.408,
|
| 640 |
+
"grad_norm": 5.446048736572266,
|
| 641 |
+
"learning_rate": 3.9918207681365574e-05,
|
| 642 |
+
"loss": 0.1013,
|
| 643 |
+
"step": 880
|
| 644 |
+
},
|
| 645 |
+
{
|
| 646 |
+
"epoch": 1.424,
|
| 647 |
+
"grad_norm": 0.017437923699617386,
|
| 648 |
+
"learning_rate": 3.9740398293029873e-05,
|
| 649 |
+
"loss": 0.0002,
|
| 650 |
+
"step": 890
|
| 651 |
+
},
|
| 652 |
+
{
|
| 653 |
+
"epoch": 1.44,
|
| 654 |
+
"grad_norm": 0.007926377467811108,
|
| 655 |
+
"learning_rate": 3.956258890469417e-05,
|
| 656 |
+
"loss": 0.0661,
|
| 657 |
+
"step": 900
|
| 658 |
+
},
|
| 659 |
+
{
|
| 660 |
+
"epoch": 1.456,
|
| 661 |
+
"grad_norm": 0.0029465279076248407,
|
| 662 |
+
"learning_rate": 3.9384779516358466e-05,
|
| 663 |
+
"loss": 0.0338,
|
| 664 |
+
"step": 910
|
| 665 |
+
},
|
| 666 |
+
{
|
| 667 |
+
"epoch": 1.472,
|
| 668 |
+
"grad_norm": 0.0010107713751494884,
|
| 669 |
+
"learning_rate": 3.9206970128022766e-05,
|
| 670 |
+
"loss": 0.0171,
|
| 671 |
+
"step": 920
|
| 672 |
+
},
|
| 673 |
+
{
|
| 674 |
+
"epoch": 1.488,
|
| 675 |
+
"grad_norm": 0.036259789019823074,
|
| 676 |
+
"learning_rate": 3.902916073968706e-05,
|
| 677 |
+
"loss": 0.0004,
|
| 678 |
+
"step": 930
|
| 679 |
+
},
|
| 680 |
+
{
|
| 681 |
+
"epoch": 1.504,
|
| 682 |
+
"grad_norm": 0.7286566495895386,
|
| 683 |
+
"learning_rate": 3.885135135135135e-05,
|
| 684 |
+
"loss": 0.014,
|
| 685 |
+
"step": 940
|
| 686 |
+
},
|
| 687 |
+
{
|
| 688 |
+
"epoch": 1.52,
|
| 689 |
+
"grad_norm": 0.008453834801912308,
|
| 690 |
+
"learning_rate": 3.8673541963015645e-05,
|
| 691 |
+
"loss": 0.0002,
|
| 692 |
+
"step": 950
|
| 693 |
+
},
|
| 694 |
+
{
|
| 695 |
+
"epoch": 1.536,
|
| 696 |
+
"grad_norm": 0.1271582841873169,
|
| 697 |
+
"learning_rate": 3.8495732574679944e-05,
|
| 698 |
+
"loss": 0.0043,
|
| 699 |
+
"step": 960
|
| 700 |
+
},
|
| 701 |
+
{
|
| 702 |
+
"epoch": 1.552,
|
| 703 |
+
"grad_norm": 6.197739124298096,
|
| 704 |
+
"learning_rate": 3.831792318634424e-05,
|
| 705 |
+
"loss": 0.0075,
|
| 706 |
+
"step": 970
|
| 707 |
+
},
|
| 708 |
+
{
|
| 709 |
+
"epoch": 1.568,
|
| 710 |
+
"grad_norm": 0.008371386677026749,
|
| 711 |
+
"learning_rate": 3.814011379800854e-05,
|
| 712 |
+
"loss": 0.0299,
|
| 713 |
+
"step": 980
|
| 714 |
+
},
|
| 715 |
+
{
|
| 716 |
+
"epoch": 1.584,
|
| 717 |
+
"grad_norm": 0.0011657042196020484,
|
| 718 |
+
"learning_rate": 3.796230440967283e-05,
|
| 719 |
+
"loss": 0.1188,
|
| 720 |
+
"step": 990
|
| 721 |
+
},
|
| 722 |
+
{
|
| 723 |
+
"epoch": 1.6,
|
| 724 |
+
"grad_norm": 0.010952652432024479,
|
| 725 |
+
"learning_rate": 3.778449502133713e-05,
|
| 726 |
+
"loss": 0.0116,
|
| 727 |
+
"step": 1000
|
| 728 |
+
},
|
| 729 |
+
{
|
| 730 |
+
"epoch": 1.616,
|
| 731 |
+
"grad_norm": 0.0010061347857117653,
|
| 732 |
+
"learning_rate": 3.760668563300143e-05,
|
| 733 |
+
"loss": 0.0043,
|
| 734 |
+
"step": 1010
|
| 735 |
+
},
|
| 736 |
+
{
|
| 737 |
+
"epoch": 1.6320000000000001,
|
| 738 |
+
"grad_norm": 0.06395132839679718,
|
| 739 |
+
"learning_rate": 3.742887624466572e-05,
|
| 740 |
+
"loss": 0.0445,
|
| 741 |
+
"step": 1020
|
| 742 |
+
},
|
| 743 |
+
{
|
| 744 |
+
"epoch": 1.6480000000000001,
|
| 745 |
+
"grad_norm": 0.001108819618821144,
|
| 746 |
+
"learning_rate": 3.7251066856330015e-05,
|
| 747 |
+
"loss": 0.0323,
|
| 748 |
+
"step": 1030
|
| 749 |
+
},
|
| 750 |
+
{
|
| 751 |
+
"epoch": 1.6640000000000001,
|
| 752 |
+
"grad_norm": 0.0014280881732702255,
|
| 753 |
+
"learning_rate": 3.707325746799431e-05,
|
| 754 |
+
"loss": 0.0008,
|
| 755 |
+
"step": 1040
|
| 756 |
+
},
|
| 757 |
+
{
|
| 758 |
+
"epoch": 1.6800000000000002,
|
| 759 |
+
"grad_norm": 0.0012123563792556524,
|
| 760 |
+
"learning_rate": 3.689544807965861e-05,
|
| 761 |
+
"loss": 0.0014,
|
| 762 |
+
"step": 1050
|
| 763 |
+
},
|
| 764 |
+
{
|
| 765 |
+
"epoch": 1.696,
|
| 766 |
+
"grad_norm": 0.0011878483928740025,
|
| 767 |
+
"learning_rate": 3.67176386913229e-05,
|
| 768 |
+
"loss": 0.0524,
|
| 769 |
+
"step": 1060
|
| 770 |
+
},
|
| 771 |
+
{
|
| 772 |
+
"epoch": 1.712,
|
| 773 |
+
"grad_norm": 0.0037567925173789263,
|
| 774 |
+
"learning_rate": 3.65398293029872e-05,
|
| 775 |
+
"loss": 0.0237,
|
| 776 |
+
"step": 1070
|
| 777 |
+
},
|
| 778 |
+
{
|
| 779 |
+
"epoch": 1.728,
|
| 780 |
+
"grad_norm": 0.002096337964758277,
|
| 781 |
+
"learning_rate": 3.636201991465149e-05,
|
| 782 |
+
"loss": 0.0334,
|
| 783 |
+
"step": 1080
|
| 784 |
+
},
|
| 785 |
+
{
|
| 786 |
+
"epoch": 1.744,
|
| 787 |
+
"grad_norm": 8.111098289489746,
|
| 788 |
+
"learning_rate": 3.618421052631579e-05,
|
| 789 |
+
"loss": 0.1467,
|
| 790 |
+
"step": 1090
|
| 791 |
+
},
|
| 792 |
+
{
|
| 793 |
+
"epoch": 1.76,
|
| 794 |
+
"grad_norm": 0.0023632964584976435,
|
| 795 |
+
"learning_rate": 3.600640113798009e-05,
|
| 796 |
+
"loss": 0.0003,
|
| 797 |
+
"step": 1100
|
| 798 |
+
},
|
| 799 |
+
{
|
| 800 |
+
"epoch": 1.776,
|
| 801 |
+
"grad_norm": 9.071378707885742,
|
| 802 |
+
"learning_rate": 3.5828591749644385e-05,
|
| 803 |
+
"loss": 0.0176,
|
| 804 |
+
"step": 1110
|
| 805 |
+
},
|
| 806 |
+
{
|
| 807 |
+
"epoch": 1.792,
|
| 808 |
+
"grad_norm": 0.08811729401350021,
|
| 809 |
+
"learning_rate": 3.565078236130868e-05,
|
| 810 |
+
"loss": 0.1454,
|
| 811 |
+
"step": 1120
|
| 812 |
+
},
|
| 813 |
+
{
|
| 814 |
+
"epoch": 1.808,
|
| 815 |
+
"grad_norm": 0.0101834237575531,
|
| 816 |
+
"learning_rate": 3.547297297297297e-05,
|
| 817 |
+
"loss": 0.0172,
|
| 818 |
+
"step": 1130
|
| 819 |
+
},
|
| 820 |
+
{
|
| 821 |
+
"epoch": 1.8239999999999998,
|
| 822 |
+
"grad_norm": 0.1676841378211975,
|
| 823 |
+
"learning_rate": 3.529516358463727e-05,
|
| 824 |
+
"loss": 0.049,
|
| 825 |
+
"step": 1140
|
| 826 |
+
},
|
| 827 |
+
{
|
| 828 |
+
"epoch": 1.8399999999999999,
|
| 829 |
+
"grad_norm": 0.043762240558862686,
|
| 830 |
+
"learning_rate": 3.5117354196301564e-05,
|
| 831 |
+
"loss": 0.0432,
|
| 832 |
+
"step": 1150
|
| 833 |
+
},
|
| 834 |
+
{
|
| 835 |
+
"epoch": 1.8559999999999999,
|
| 836 |
+
"grad_norm": 0.011058449745178223,
|
| 837 |
+
"learning_rate": 3.4939544807965863e-05,
|
| 838 |
+
"loss": 0.001,
|
| 839 |
+
"step": 1160
|
| 840 |
+
},
|
| 841 |
+
{
|
| 842 |
+
"epoch": 1.8719999999999999,
|
| 843 |
+
"grad_norm": 0.0018902173032984138,
|
| 844 |
+
"learning_rate": 3.4761735419630156e-05,
|
| 845 |
+
"loss": 0.0399,
|
| 846 |
+
"step": 1170
|
| 847 |
+
},
|
| 848 |
+
{
|
| 849 |
+
"epoch": 1.888,
|
| 850 |
+
"grad_norm": 0.001800389145500958,
|
| 851 |
+
"learning_rate": 3.4583926031294456e-05,
|
| 852 |
+
"loss": 0.0184,
|
| 853 |
+
"step": 1180
|
| 854 |
+
},
|
| 855 |
+
{
|
| 856 |
+
"epoch": 1.904,
|
| 857 |
+
"grad_norm": 0.06505569815635681,
|
| 858 |
+
"learning_rate": 3.440611664295875e-05,
|
| 859 |
+
"loss": 0.0073,
|
| 860 |
+
"step": 1190
|
| 861 |
+
},
|
| 862 |
+
{
|
| 863 |
+
"epoch": 1.92,
|
| 864 |
+
"grad_norm": 0.5011488199234009,
|
| 865 |
+
"learning_rate": 3.422830725462304e-05,
|
| 866 |
+
"loss": 0.0007,
|
| 867 |
+
"step": 1200
|
| 868 |
+
},
|
| 869 |
+
{
|
| 870 |
+
"epoch": 1.936,
|
| 871 |
+
"grad_norm": 0.013254979625344276,
|
| 872 |
+
"learning_rate": 3.405049786628734e-05,
|
| 873 |
+
"loss": 0.0059,
|
| 874 |
+
"step": 1210
|
| 875 |
+
},
|
| 876 |
+
{
|
| 877 |
+
"epoch": 1.952,
|
| 878 |
+
"grad_norm": 0.7582728266716003,
|
| 879 |
+
"learning_rate": 3.3872688477951634e-05,
|
| 880 |
+
"loss": 0.0013,
|
| 881 |
+
"step": 1220
|
| 882 |
+
},
|
| 883 |
+
{
|
| 884 |
+
"epoch": 1.968,
|
| 885 |
+
"grad_norm": 0.009909105487167835,
|
| 886 |
+
"learning_rate": 3.3694879089615934e-05,
|
| 887 |
+
"loss": 0.0002,
|
| 888 |
+
"step": 1230
|
| 889 |
+
},
|
| 890 |
+
{
|
| 891 |
+
"epoch": 1.984,
|
| 892 |
+
"grad_norm": 0.000718823226634413,
|
| 893 |
+
"learning_rate": 3.351706970128023e-05,
|
| 894 |
+
"loss": 0.0047,
|
| 895 |
+
"step": 1240
|
| 896 |
+
},
|
| 897 |
+
{
|
| 898 |
+
"epoch": 2.0,
|
| 899 |
+
"grad_norm": 0.6642739772796631,
|
| 900 |
+
"learning_rate": 3.333926031294453e-05,
|
| 901 |
+
"loss": 0.0015,
|
| 902 |
+
"step": 1250
|
| 903 |
+
},
|
| 904 |
+
{
|
| 905 |
+
"epoch": 2.0,
|
| 906 |
+
"eval_accuracy": 0.9919839679358717,
|
| 907 |
+
"eval_f1_macro": 0.9810672282141238,
|
| 908 |
+
"eval_f1_micro": 0.9919839679358717,
|
| 909 |
+
"eval_f1_weighted": 0.992107547860045,
|
| 910 |
+
"eval_loss": 0.019453825429081917,
|
| 911 |
+
"eval_precision_macro": 0.9692164931816348,
|
| 912 |
+
"eval_precision_micro": 0.9919839679358717,
|
| 913 |
+
"eval_precision_weighted": 0.9925061936994298,
|
| 914 |
+
"eval_recall_macro": 0.9942915690866512,
|
| 915 |
+
"eval_recall_micro": 0.9919839679358717,
|
| 916 |
+
"eval_recall_weighted": 0.9919839679358717,
|
| 917 |
+
"eval_runtime": 10.636,
|
| 918 |
+
"eval_samples_per_second": 46.916,
|
| 919 |
+
"eval_steps_per_second": 3.009,
|
| 920 |
+
"step": 1250
|
| 921 |
+
},
|
| 922 |
+
{
|
| 923 |
+
"epoch": 2.016,
|
| 924 |
+
"grad_norm": 0.005395730957388878,
|
| 925 |
+
"learning_rate": 3.316145092460882e-05,
|
| 926 |
+
"loss": 0.0001,
|
| 927 |
+
"step": 1260
|
| 928 |
+
},
|
| 929 |
+
{
|
| 930 |
+
"epoch": 2.032,
|
| 931 |
+
"grad_norm": 0.0008304046932607889,
|
| 932 |
+
"learning_rate": 3.298364153627312e-05,
|
| 933 |
+
"loss": 0.0001,
|
| 934 |
+
"step": 1270
|
| 935 |
+
},
|
| 936 |
+
{
|
| 937 |
+
"epoch": 2.048,
|
| 938 |
+
"grad_norm": 0.0007791437674313784,
|
| 939 |
+
"learning_rate": 3.280583214793741e-05,
|
| 940 |
+
"loss": 0.0001,
|
| 941 |
+
"step": 1280
|
| 942 |
+
},
|
| 943 |
+
{
|
| 944 |
+
"epoch": 2.064,
|
| 945 |
+
"grad_norm": 0.0008820474613457918,
|
| 946 |
+
"learning_rate": 3.2628022759601705e-05,
|
| 947 |
+
"loss": 0.0001,
|
| 948 |
+
"step": 1290
|
| 949 |
+
},
|
| 950 |
+
{
|
| 951 |
+
"epoch": 2.08,
|
| 952 |
+
"grad_norm": 0.000594152370467782,
|
| 953 |
+
"learning_rate": 3.2450213371266005e-05,
|
| 954 |
+
"loss": 0.0002,
|
| 955 |
+
"step": 1300
|
| 956 |
+
},
|
| 957 |
+
{
|
| 958 |
+
"epoch": 2.096,
|
| 959 |
+
"grad_norm": 0.0006243674433790147,
|
| 960 |
+
"learning_rate": 3.22724039829303e-05,
|
| 961 |
+
"loss": 0.0022,
|
| 962 |
+
"step": 1310
|
| 963 |
+
},
|
| 964 |
+
{
|
| 965 |
+
"epoch": 2.112,
|
| 966 |
+
"grad_norm": 0.0007709413184784353,
|
| 967 |
+
"learning_rate": 3.20945945945946e-05,
|
| 968 |
+
"loss": 0.0154,
|
| 969 |
+
"step": 1320
|
| 970 |
+
},
|
| 971 |
+
{
|
| 972 |
+
"epoch": 2.128,
|
| 973 |
+
"grad_norm": 0.004626740701496601,
|
| 974 |
+
"learning_rate": 3.191678520625889e-05,
|
| 975 |
+
"loss": 0.0012,
|
| 976 |
+
"step": 1330
|
| 977 |
+
},
|
| 978 |
+
{
|
| 979 |
+
"epoch": 2.144,
|
| 980 |
+
"grad_norm": 0.002739989897236228,
|
| 981 |
+
"learning_rate": 3.173897581792319e-05,
|
| 982 |
+
"loss": 0.0004,
|
| 983 |
+
"step": 1340
|
| 984 |
+
},
|
| 985 |
+
{
|
| 986 |
+
"epoch": 2.16,
|
| 987 |
+
"grad_norm": 0.007507917005568743,
|
| 988 |
+
"learning_rate": 3.156116642958748e-05,
|
| 989 |
+
"loss": 0.0001,
|
| 990 |
+
"step": 1350
|
| 991 |
+
},
|
| 992 |
+
{
|
| 993 |
+
"epoch": 2.176,
|
| 994 |
+
"grad_norm": 0.004014975391328335,
|
| 995 |
+
"learning_rate": 3.138335704125178e-05,
|
| 996 |
+
"loss": 0.001,
|
| 997 |
+
"step": 1360
|
| 998 |
+
},
|
| 999 |
+
{
|
| 1000 |
+
"epoch": 2.192,
|
| 1001 |
+
"grad_norm": 0.000455196452094242,
|
| 1002 |
+
"learning_rate": 3.1205547652916076e-05,
|
| 1003 |
+
"loss": 0.0005,
|
| 1004 |
+
"step": 1370
|
| 1005 |
+
},
|
| 1006 |
+
{
|
| 1007 |
+
"epoch": 2.208,
|
| 1008 |
+
"grad_norm": 0.00030664558289572597,
|
| 1009 |
+
"learning_rate": 3.102773826458037e-05,
|
| 1010 |
+
"loss": 0.0265,
|
| 1011 |
+
"step": 1380
|
| 1012 |
+
},
|
| 1013 |
+
{
|
| 1014 |
+
"epoch": 2.224,
|
| 1015 |
+
"grad_norm": 0.00043227567221038043,
|
| 1016 |
+
"learning_rate": 3.084992887624467e-05,
|
| 1017 |
+
"loss": 0.0001,
|
| 1018 |
+
"step": 1390
|
| 1019 |
+
},
|
| 1020 |
+
{
|
| 1021 |
+
"epoch": 2.24,
|
| 1022 |
+
"grad_norm": 0.0481143593788147,
|
| 1023 |
+
"learning_rate": 3.067211948790896e-05,
|
| 1024 |
+
"loss": 0.0023,
|
| 1025 |
+
"step": 1400
|
| 1026 |
+
},
|
| 1027 |
+
{
|
| 1028 |
+
"epoch": 2.2560000000000002,
|
| 1029 |
+
"grad_norm": 0.005968974903225899,
|
| 1030 |
+
"learning_rate": 3.0494310099573257e-05,
|
| 1031 |
+
"loss": 0.0025,
|
| 1032 |
+
"step": 1410
|
| 1033 |
+
},
|
| 1034 |
+
{
|
| 1035 |
+
"epoch": 2.2720000000000002,
|
| 1036 |
+
"grad_norm": 0.21466164290905,
|
| 1037 |
+
"learning_rate": 3.0316500711237557e-05,
|
| 1038 |
+
"loss": 0.0003,
|
| 1039 |
+
"step": 1420
|
| 1040 |
+
},
|
| 1041 |
+
{
|
| 1042 |
+
"epoch": 2.288,
|
| 1043 |
+
"grad_norm": 0.0018357799854129553,
|
| 1044 |
+
"learning_rate": 3.0138691322901853e-05,
|
| 1045 |
+
"loss": 0.0001,
|
| 1046 |
+
"step": 1430
|
| 1047 |
+
},
|
| 1048 |
+
{
|
| 1049 |
+
"epoch": 2.304,
|
| 1050 |
+
"grad_norm": 0.0004517412162385881,
|
| 1051 |
+
"learning_rate": 2.996088193456615e-05,
|
| 1052 |
+
"loss": 0.0001,
|
| 1053 |
+
"step": 1440
|
| 1054 |
+
},
|
| 1055 |
+
{
|
| 1056 |
+
"epoch": 2.32,
|
| 1057 |
+
"grad_norm": 0.005384071730077267,
|
| 1058 |
+
"learning_rate": 2.9783072546230446e-05,
|
| 1059 |
+
"loss": 0.002,
|
| 1060 |
+
"step": 1450
|
| 1061 |
+
},
|
| 1062 |
+
{
|
| 1063 |
+
"epoch": 2.336,
|
| 1064 |
+
"grad_norm": 0.00046941509936004877,
|
| 1065 |
+
"learning_rate": 2.9605263157894735e-05,
|
| 1066 |
+
"loss": 0.0001,
|
| 1067 |
+
"step": 1460
|
| 1068 |
+
},
|
| 1069 |
+
{
|
| 1070 |
+
"epoch": 2.352,
|
| 1071 |
+
"grad_norm": 0.011709867045283318,
|
| 1072 |
+
"learning_rate": 2.9427453769559032e-05,
|
| 1073 |
+
"loss": 0.0001,
|
| 1074 |
+
"step": 1470
|
| 1075 |
+
},
|
| 1076 |
+
{
|
| 1077 |
+
"epoch": 2.368,
|
| 1078 |
+
"grad_norm": 0.03616934269666672,
|
| 1079 |
+
"learning_rate": 2.9249644381223328e-05,
|
| 1080 |
+
"loss": 0.0008,
|
| 1081 |
+
"step": 1480
|
| 1082 |
+
},
|
| 1083 |
+
{
|
| 1084 |
+
"epoch": 2.384,
|
| 1085 |
+
"grad_norm": 0.0006385542219504714,
|
| 1086 |
+
"learning_rate": 2.9071834992887624e-05,
|
| 1087 |
+
"loss": 0.0001,
|
| 1088 |
+
"step": 1490
|
| 1089 |
+
},
|
| 1090 |
+
{
|
| 1091 |
+
"epoch": 2.4,
|
| 1092 |
+
"grad_norm": 1.0423965454101562,
|
| 1093 |
+
"learning_rate": 2.889402560455192e-05,
|
| 1094 |
+
"loss": 0.0021,
|
| 1095 |
+
"step": 1500
|
| 1096 |
+
},
|
| 1097 |
+
{
|
| 1098 |
+
"epoch": 2.416,
|
| 1099 |
+
"grad_norm": 0.21064622700214386,
|
| 1100 |
+
"learning_rate": 2.8716216216216217e-05,
|
| 1101 |
+
"loss": 0.0003,
|
| 1102 |
+
"step": 1510
|
| 1103 |
+
},
|
| 1104 |
+
{
|
| 1105 |
+
"epoch": 2.432,
|
| 1106 |
+
"grad_norm": 0.0016910170670598745,
|
| 1107 |
+
"learning_rate": 2.8538406827880517e-05,
|
| 1108 |
+
"loss": 0.0001,
|
| 1109 |
+
"step": 1520
|
| 1110 |
+
},
|
| 1111 |
+
{
|
| 1112 |
+
"epoch": 2.448,
|
| 1113 |
+
"grad_norm": 0.0004065225657541305,
|
| 1114 |
+
"learning_rate": 2.8360597439544813e-05,
|
| 1115 |
+
"loss": 0.0004,
|
| 1116 |
+
"step": 1530
|
| 1117 |
+
},
|
| 1118 |
+
{
|
| 1119 |
+
"epoch": 2.464,
|
| 1120 |
+
"grad_norm": 0.0006998078897595406,
|
| 1121 |
+
"learning_rate": 2.8182788051209103e-05,
|
| 1122 |
+
"loss": 0.0001,
|
| 1123 |
+
"step": 1540
|
| 1124 |
+
},
|
| 1125 |
+
{
|
| 1126 |
+
"epoch": 2.48,
|
| 1127 |
+
"grad_norm": 0.0057923863641917706,
|
| 1128 |
+
"learning_rate": 2.80049786628734e-05,
|
| 1129 |
+
"loss": 0.0012,
|
| 1130 |
+
"step": 1550
|
| 1131 |
+
},
|
| 1132 |
+
{
|
| 1133 |
+
"epoch": 2.496,
|
| 1134 |
+
"grad_norm": 0.00030601295293308794,
|
| 1135 |
+
"learning_rate": 2.7827169274537695e-05,
|
| 1136 |
+
"loss": 0.0236,
|
| 1137 |
+
"step": 1560
|
| 1138 |
+
},
|
| 1139 |
+
{
|
| 1140 |
+
"epoch": 2.512,
|
| 1141 |
+
"grad_norm": 0.00045617681462317705,
|
| 1142 |
+
"learning_rate": 2.764935988620199e-05,
|
| 1143 |
+
"loss": 0.0001,
|
| 1144 |
+
"step": 1570
|
| 1145 |
+
},
|
| 1146 |
+
{
|
| 1147 |
+
"epoch": 2.528,
|
| 1148 |
+
"grad_norm": 0.00044945545960217714,
|
| 1149 |
+
"learning_rate": 2.7471550497866288e-05,
|
| 1150 |
+
"loss": 0.0001,
|
| 1151 |
+
"step": 1580
|
| 1152 |
+
},
|
| 1153 |
+
{
|
| 1154 |
+
"epoch": 2.544,
|
| 1155 |
+
"grad_norm": 0.0003521046892274171,
|
| 1156 |
+
"learning_rate": 2.7293741109530584e-05,
|
| 1157 |
+
"loss": 0.0263,
|
| 1158 |
+
"step": 1590
|
| 1159 |
+
},
|
| 1160 |
+
{
|
| 1161 |
+
"epoch": 2.56,
|
| 1162 |
+
"grad_norm": 0.008946732617914677,
|
| 1163 |
+
"learning_rate": 2.711593172119488e-05,
|
| 1164 |
+
"loss": 0.0006,
|
| 1165 |
+
"step": 1600
|
| 1166 |
+
},
|
| 1167 |
+
{
|
| 1168 |
+
"epoch": 2.576,
|
| 1169 |
+
"grad_norm": 0.09531024098396301,
|
| 1170 |
+
"learning_rate": 2.6938122332859177e-05,
|
| 1171 |
+
"loss": 0.0003,
|
| 1172 |
+
"step": 1610
|
| 1173 |
+
},
|
| 1174 |
+
{
|
| 1175 |
+
"epoch": 2.592,
|
| 1176 |
+
"grad_norm": 0.0010131685994565487,
|
| 1177 |
+
"learning_rate": 2.6760312944523473e-05,
|
| 1178 |
+
"loss": 0.0001,
|
| 1179 |
+
"step": 1620
|
| 1180 |
+
},
|
| 1181 |
+
{
|
| 1182 |
+
"epoch": 2.608,
|
| 1183 |
+
"grad_norm": 0.0016723967855796218,
|
| 1184 |
+
"learning_rate": 2.6582503556187766e-05,
|
| 1185 |
+
"loss": 0.0018,
|
| 1186 |
+
"step": 1630
|
| 1187 |
+
},
|
| 1188 |
+
{
|
| 1189 |
+
"epoch": 2.624,
|
| 1190 |
+
"grad_norm": 0.008407480083405972,
|
| 1191 |
+
"learning_rate": 2.6404694167852062e-05,
|
| 1192 |
+
"loss": 0.0344,
|
| 1193 |
+
"step": 1640
|
| 1194 |
+
},
|
| 1195 |
+
{
|
| 1196 |
+
"epoch": 2.64,
|
| 1197 |
+
"grad_norm": 0.001671052654273808,
|
| 1198 |
+
"learning_rate": 2.622688477951636e-05,
|
| 1199 |
+
"loss": 0.0005,
|
| 1200 |
+
"step": 1650
|
| 1201 |
+
},
|
| 1202 |
+
{
|
| 1203 |
+
"epoch": 2.656,
|
| 1204 |
+
"grad_norm": 0.0003514339041430503,
|
| 1205 |
+
"learning_rate": 2.6049075391180655e-05,
|
| 1206 |
+
"loss": 0.0001,
|
| 1207 |
+
"step": 1660
|
| 1208 |
+
},
|
| 1209 |
+
{
|
| 1210 |
+
"epoch": 2.672,
|
| 1211 |
+
"grad_norm": 0.00044292688835412264,
|
| 1212 |
+
"learning_rate": 2.587126600284495e-05,
|
| 1213 |
+
"loss": 0.0017,
|
| 1214 |
+
"step": 1670
|
| 1215 |
+
},
|
| 1216 |
+
{
|
| 1217 |
+
"epoch": 2.6879999999999997,
|
| 1218 |
+
"grad_norm": 0.00033179231104440987,
|
| 1219 |
+
"learning_rate": 2.5693456614509247e-05,
|
| 1220 |
+
"loss": 0.0243,
|
| 1221 |
+
"step": 1680
|
| 1222 |
+
},
|
| 1223 |
+
{
|
| 1224 |
+
"epoch": 2.7039999999999997,
|
| 1225 |
+
"grad_norm": 0.0023233199026435614,
|
| 1226 |
+
"learning_rate": 2.5515647226173544e-05,
|
| 1227 |
+
"loss": 0.0001,
|
| 1228 |
+
"step": 1690
|
| 1229 |
+
},
|
| 1230 |
+
{
|
| 1231 |
+
"epoch": 2.7199999999999998,
|
| 1232 |
+
"grad_norm": 0.0012358427047729492,
|
| 1233 |
+
"learning_rate": 2.533783783783784e-05,
|
| 1234 |
+
"loss": 0.0002,
|
| 1235 |
+
"step": 1700
|
| 1236 |
+
},
|
| 1237 |
+
{
|
| 1238 |
+
"epoch": 2.7359999999999998,
|
| 1239 |
+
"grad_norm": 0.00035215960815548897,
|
| 1240 |
+
"learning_rate": 2.5160028449502136e-05,
|
| 1241 |
+
"loss": 0.0235,
|
| 1242 |
+
"step": 1710
|
| 1243 |
+
},
|
| 1244 |
+
{
|
| 1245 |
+
"epoch": 2.752,
|
| 1246 |
+
"grad_norm": 0.24424146115779877,
|
| 1247 |
+
"learning_rate": 2.4982219061166433e-05,
|
| 1248 |
+
"loss": 0.0005,
|
| 1249 |
+
"step": 1720
|
| 1250 |
+
},
|
| 1251 |
+
{
|
| 1252 |
+
"epoch": 2.768,
|
| 1253 |
+
"grad_norm": 2.014695167541504,
|
| 1254 |
+
"learning_rate": 2.480440967283073e-05,
|
| 1255 |
+
"loss": 0.0111,
|
| 1256 |
+
"step": 1730
|
| 1257 |
+
},
|
| 1258 |
+
{
|
| 1259 |
+
"epoch": 2.784,
|
| 1260 |
+
"grad_norm": 0.00039951372309587896,
|
| 1261 |
+
"learning_rate": 2.4626600284495022e-05,
|
| 1262 |
+
"loss": 0.0004,
|
| 1263 |
+
"step": 1740
|
| 1264 |
+
},
|
| 1265 |
+
{
|
| 1266 |
+
"epoch": 2.8,
|
| 1267 |
+
"grad_norm": 0.0016597098438069224,
|
| 1268 |
+
"learning_rate": 2.4448790896159318e-05,
|
| 1269 |
+
"loss": 0.0002,
|
| 1270 |
+
"step": 1750
|
| 1271 |
+
},
|
| 1272 |
+
{
|
| 1273 |
+
"epoch": 2.816,
|
| 1274 |
+
"grad_norm": 0.010242385789752007,
|
| 1275 |
+
"learning_rate": 2.4270981507823614e-05,
|
| 1276 |
+
"loss": 0.0001,
|
| 1277 |
+
"step": 1760
|
| 1278 |
+
},
|
| 1279 |
+
{
|
| 1280 |
+
"epoch": 2.832,
|
| 1281 |
+
"grad_norm": 0.0023806928656995296,
|
| 1282 |
+
"learning_rate": 2.409317211948791e-05,
|
| 1283 |
+
"loss": 0.0001,
|
| 1284 |
+
"step": 1770
|
| 1285 |
+
},
|
| 1286 |
+
{
|
| 1287 |
+
"epoch": 2.848,
|
| 1288 |
+
"grad_norm": 0.01970355026423931,
|
| 1289 |
+
"learning_rate": 2.3915362731152204e-05,
|
| 1290 |
+
"loss": 0.0183,
|
| 1291 |
+
"step": 1780
|
| 1292 |
+
},
|
| 1293 |
+
{
|
| 1294 |
+
"epoch": 2.864,
|
| 1295 |
+
"grad_norm": 0.0002463227428961545,
|
| 1296 |
+
"learning_rate": 2.37375533428165e-05,
|
| 1297 |
+
"loss": 0.0,
|
| 1298 |
+
"step": 1790
|
| 1299 |
+
},
|
| 1300 |
+
{
|
| 1301 |
+
"epoch": 2.88,
|
| 1302 |
+
"grad_norm": 0.000822290952783078,
|
| 1303 |
+
"learning_rate": 2.35597439544808e-05,
|
| 1304 |
+
"loss": 0.0001,
|
| 1305 |
+
"step": 1800
|
| 1306 |
+
},
|
| 1307 |
+
{
|
| 1308 |
+
"epoch": 2.896,
|
| 1309 |
+
"grad_norm": 0.0026505696587264538,
|
| 1310 |
+
"learning_rate": 2.3381934566145096e-05,
|
| 1311 |
+
"loss": 0.0004,
|
| 1312 |
+
"step": 1810
|
| 1313 |
+
},
|
| 1314 |
+
{
|
| 1315 |
+
"epoch": 2.912,
|
| 1316 |
+
"grad_norm": 0.0005039023817516863,
|
| 1317 |
+
"learning_rate": 2.320412517780939e-05,
|
| 1318 |
+
"loss": 0.0001,
|
| 1319 |
+
"step": 1820
|
| 1320 |
+
},
|
| 1321 |
+
{
|
| 1322 |
+
"epoch": 2.928,
|
| 1323 |
+
"grad_norm": 0.009028232656419277,
|
| 1324 |
+
"learning_rate": 2.3026315789473685e-05,
|
| 1325 |
+
"loss": 0.0001,
|
| 1326 |
+
"step": 1830
|
| 1327 |
+
},
|
| 1328 |
+
{
|
| 1329 |
+
"epoch": 2.944,
|
| 1330 |
+
"grad_norm": 0.002705842722207308,
|
| 1331 |
+
"learning_rate": 2.284850640113798e-05,
|
| 1332 |
+
"loss": 0.0003,
|
| 1333 |
+
"step": 1840
|
| 1334 |
+
},
|
| 1335 |
+
{
|
| 1336 |
+
"epoch": 2.96,
|
| 1337 |
+
"grad_norm": 0.00383372837677598,
|
| 1338 |
+
"learning_rate": 2.2670697012802278e-05,
|
| 1339 |
+
"loss": 0.0017,
|
| 1340 |
+
"step": 1850
|
| 1341 |
+
},
|
| 1342 |
+
{
|
| 1343 |
+
"epoch": 2.976,
|
| 1344 |
+
"grad_norm": 0.00024894202942959964,
|
| 1345 |
+
"learning_rate": 2.2492887624466574e-05,
|
| 1346 |
+
"loss": 0.0002,
|
| 1347 |
+
"step": 1860
|
| 1348 |
+
},
|
| 1349 |
+
{
|
| 1350 |
+
"epoch": 2.992,
|
| 1351 |
+
"grad_norm": 0.0009176091407425702,
|
| 1352 |
+
"learning_rate": 2.2315078236130867e-05,
|
| 1353 |
+
"loss": 0.0,
|
| 1354 |
+
"step": 1870
|
| 1355 |
+
},
|
| 1356 |
+
{
|
| 1357 |
+
"epoch": 3.0,
|
| 1358 |
+
"eval_accuracy": 0.9939879759519038,
|
| 1359 |
+
"eval_f1_macro": 0.9910040909369707,
|
| 1360 |
+
"eval_f1_micro": 0.9939879759519038,
|
| 1361 |
+
"eval_f1_weighted": 0.9939809297459037,
|
| 1362 |
+
"eval_loss": 0.014622284099459648,
|
| 1363 |
+
"eval_precision_macro": 0.9959514170040485,
|
| 1364 |
+
"eval_precision_micro": 0.9939879759519038,
|
| 1365 |
+
"eval_precision_weighted": 0.994060996486901,
|
| 1366 |
+
"eval_recall_macro": 0.9862711213517666,
|
| 1367 |
+
"eval_recall_micro": 0.9939879759519038,
|
| 1368 |
+
"eval_recall_weighted": 0.9939879759519038,
|
| 1369 |
+
"eval_runtime": 10.5217,
|
| 1370 |
+
"eval_samples_per_second": 47.426,
|
| 1371 |
+
"eval_steps_per_second": 3.041,
|
| 1372 |
+
"step": 1875
|
| 1373 |
+
}
|
| 1374 |
+
],
|
| 1375 |
+
"logging_steps": 10,
|
| 1376 |
+
"max_steps": 3125,
|
| 1377 |
+
"num_input_tokens_seen": 0,
|
| 1378 |
+
"num_train_epochs": 5,
|
| 1379 |
+
"save_steps": 500,
|
| 1380 |
+
"stateful_callbacks": {
|
| 1381 |
+
"EarlyStoppingCallback": {
|
| 1382 |
+
"args": {
|
| 1383 |
+
"early_stopping_patience": 5,
|
| 1384 |
+
"early_stopping_threshold": 0.01
|
| 1385 |
+
},
|
| 1386 |
+
"attributes": {
|
| 1387 |
+
"early_stopping_patience_counter": 2
|
| 1388 |
+
}
|
| 1389 |
+
},
|
| 1390 |
+
"TrainerControl": {
|
| 1391 |
+
"args": {
|
| 1392 |
+
"should_epoch_stop": false,
|
| 1393 |
+
"should_evaluate": false,
|
| 1394 |
+
"should_log": false,
|
| 1395 |
+
"should_save": true,
|
| 1396 |
+
"should_training_stop": false
|
| 1397 |
+
},
|
| 1398 |
+
"attributes": {}
|
| 1399 |
+
}
|
| 1400 |
+
},
|
| 1401 |
+
"total_flos": 2446967439360000.0,
|
| 1402 |
+
"train_batch_size": 8,
|
| 1403 |
+
"trial_name": null,
|
| 1404 |
+
"trial_params": null
|
| 1405 |
+
}
|
training_args.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:3ea603352d747c992629b055015f8d059765cdd964379a0825a7adeea2eb9304
|
| 3 |
+
size 5777
|