Upload 8 files
Browse files- config.json +59 -0
- model.safetensors +3 -0
- optimizer.pt +3 -0
- rng_state.pth +3 -0
- scaler.pt +3 -0
- scheduler.pt +3 -0
- trainer_state.json +1699 -0
- training_args.bin +3 -0
config.json
ADDED
|
@@ -0,0 +1,59 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"_name_or_path": "openai-community/gpt2",
|
| 3 |
+
"activation_function": "gelu_new",
|
| 4 |
+
"architectures": [
|
| 5 |
+
"GPT2ForSequenceClassification"
|
| 6 |
+
],
|
| 7 |
+
"attn_pdrop": 0.1,
|
| 8 |
+
"bos_token_id": 50256,
|
| 9 |
+
"embd_pdrop": 0.1,
|
| 10 |
+
"eos_token_id": 50256,
|
| 11 |
+
"id2label": {
|
| 12 |
+
"0": "LABEL_0",
|
| 13 |
+
"1": "LABEL_1",
|
| 14 |
+
"2": "LABEL_2",
|
| 15 |
+
"3": "LABEL_3",
|
| 16 |
+
"4": "LABEL_4",
|
| 17 |
+
"5": "LABEL_5",
|
| 18 |
+
"6": "LABEL_6"
|
| 19 |
+
},
|
| 20 |
+
"initializer_range": 0.02,
|
| 21 |
+
"label2id": {
|
| 22 |
+
"LABEL_0": 0,
|
| 23 |
+
"LABEL_1": 1,
|
| 24 |
+
"LABEL_2": 2,
|
| 25 |
+
"LABEL_3": 3,
|
| 26 |
+
"LABEL_4": 4,
|
| 27 |
+
"LABEL_5": 5,
|
| 28 |
+
"LABEL_6": 6
|
| 29 |
+
},
|
| 30 |
+
"layer_norm_epsilon": 1e-05,
|
| 31 |
+
"model_type": "gpt2",
|
| 32 |
+
"n_ctx": 1024,
|
| 33 |
+
"n_embd": 768,
|
| 34 |
+
"n_head": 12,
|
| 35 |
+
"n_inner": null,
|
| 36 |
+
"n_layer": 12,
|
| 37 |
+
"n_positions": 1024,
|
| 38 |
+
"pad_token_id": 50256,
|
| 39 |
+
"problem_type": "single_label_classification",
|
| 40 |
+
"reorder_and_upcast_attn": false,
|
| 41 |
+
"resid_pdrop": 0.1,
|
| 42 |
+
"scale_attn_by_inverse_layer_idx": false,
|
| 43 |
+
"scale_attn_weights": true,
|
| 44 |
+
"summary_activation": null,
|
| 45 |
+
"summary_first_dropout": 0.1,
|
| 46 |
+
"summary_proj_to_labels": true,
|
| 47 |
+
"summary_type": "cls_index",
|
| 48 |
+
"summary_use_proj": true,
|
| 49 |
+
"task_specific_params": {
|
| 50 |
+
"text-generation": {
|
| 51 |
+
"do_sample": true,
|
| 52 |
+
"max_length": 50
|
| 53 |
+
}
|
| 54 |
+
},
|
| 55 |
+
"torch_dtype": "float32",
|
| 56 |
+
"transformers_version": "4.49.0",
|
| 57 |
+
"use_cache": true,
|
| 58 |
+
"vocab_size": 50257
|
| 59 |
+
}
|
model.safetensors
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:d0fc38abd2423000deac57ae4a3d8817ff4bc2330d719a447f94ccd95810b6bc
|
| 3 |
+
size 497795792
|
optimizer.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:a0f69c0da68472420420e7035f0117250e64b5a441a109d859637922cf92ab64
|
| 3 |
+
size 995686138
|
rng_state.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:186f463110bb3b61f265e81b827eb6cfe9e9853b320a88016808d64284d272d7
|
| 3 |
+
size 14244
|
scaler.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:05fdd26cf6851b130b583e49ff5643be0d0a64e3b5c5f62158e5106f14e6144d
|
| 3 |
+
size 988
|
scheduler.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:8ab94b495a347b0fdc27086b8b7987c756563e833856d443bc82b8ca4b65f4d1
|
| 3 |
+
size 1064
|
trainer_state.json
ADDED
|
@@ -0,0 +1,1699 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": 1.7819397449493408,
|
| 3 |
+
"best_model_checkpoint": ".cache/models/de3349474c859e664bc22aa79edda407/checkpoint-116484",
|
| 4 |
+
"epoch": 3.0,
|
| 5 |
+
"eval_steps": 500,
|
| 6 |
+
"global_step": 116484,
|
| 7 |
+
"is_hyper_param_search": false,
|
| 8 |
+
"is_local_process_zero": true,
|
| 9 |
+
"is_world_process_zero": true,
|
| 10 |
+
"log_history": [
|
| 11 |
+
{
|
| 12 |
+
"epoch": 0.012877305037601731,
|
| 13 |
+
"grad_norm": 9.567660331726074,
|
| 14 |
+
"learning_rate": 6.998222416812609e-06,
|
| 15 |
+
"loss": 2.1235,
|
| 16 |
+
"step": 500
|
| 17 |
+
},
|
| 18 |
+
{
|
| 19 |
+
"epoch": 0.025754610075203462,
|
| 20 |
+
"grad_norm": 8.706146240234375,
|
| 21 |
+
"learning_rate": 6.996419594107345e-06,
|
| 22 |
+
"loss": 1.9258,
|
| 23 |
+
"step": 1000
|
| 24 |
+
},
|
| 25 |
+
{
|
| 26 |
+
"epoch": 0.03863191511280519,
|
| 27 |
+
"grad_norm": 11.026541709899902,
|
| 28 |
+
"learning_rate": 6.9946167714020805e-06,
|
| 29 |
+
"loss": 1.8954,
|
| 30 |
+
"step": 1500
|
| 31 |
+
},
|
| 32 |
+
{
|
| 33 |
+
"epoch": 0.051509220150406924,
|
| 34 |
+
"grad_norm": 11.393457412719727,
|
| 35 |
+
"learning_rate": 6.992813948696817e-06,
|
| 36 |
+
"loss": 1.8783,
|
| 37 |
+
"step": 2000
|
| 38 |
+
},
|
| 39 |
+
{
|
| 40 |
+
"epoch": 0.06438652518800865,
|
| 41 |
+
"grad_norm": 6.881910800933838,
|
| 42 |
+
"learning_rate": 6.991011125991553e-06,
|
| 43 |
+
"loss": 1.8627,
|
| 44 |
+
"step": 2500
|
| 45 |
+
},
|
| 46 |
+
{
|
| 47 |
+
"epoch": 0.07726383022561038,
|
| 48 |
+
"grad_norm": 9.6551513671875,
|
| 49 |
+
"learning_rate": 6.989208303286289e-06,
|
| 50 |
+
"loss": 1.8543,
|
| 51 |
+
"step": 3000
|
| 52 |
+
},
|
| 53 |
+
{
|
| 54 |
+
"epoch": 0.09014113526321212,
|
| 55 |
+
"grad_norm": 7.362414836883545,
|
| 56 |
+
"learning_rate": 6.987405480581024e-06,
|
| 57 |
+
"loss": 1.825,
|
| 58 |
+
"step": 3500
|
| 59 |
+
},
|
| 60 |
+
{
|
| 61 |
+
"epoch": 0.10301844030081385,
|
| 62 |
+
"grad_norm": 7.49143123626709,
|
| 63 |
+
"learning_rate": 6.9856026578757595e-06,
|
| 64 |
+
"loss": 1.822,
|
| 65 |
+
"step": 4000
|
| 66 |
+
},
|
| 67 |
+
{
|
| 68 |
+
"epoch": 0.11589574533841558,
|
| 69 |
+
"grad_norm": 5.751263618469238,
|
| 70 |
+
"learning_rate": 6.983799835170495e-06,
|
| 71 |
+
"loss": 1.8325,
|
| 72 |
+
"step": 4500
|
| 73 |
+
},
|
| 74 |
+
{
|
| 75 |
+
"epoch": 0.1287730503760173,
|
| 76 |
+
"grad_norm": 12.193325996398926,
|
| 77 |
+
"learning_rate": 6.981997012465231e-06,
|
| 78 |
+
"loss": 1.8237,
|
| 79 |
+
"step": 5000
|
| 80 |
+
},
|
| 81 |
+
{
|
| 82 |
+
"epoch": 0.14165035541361903,
|
| 83 |
+
"grad_norm": 5.82090950012207,
|
| 84 |
+
"learning_rate": 6.980194189759967e-06,
|
| 85 |
+
"loss": 1.8273,
|
| 86 |
+
"step": 5500
|
| 87 |
+
},
|
| 88 |
+
{
|
| 89 |
+
"epoch": 0.15452766045122077,
|
| 90 |
+
"grad_norm": 10.159476280212402,
|
| 91 |
+
"learning_rate": 6.978391367054702e-06,
|
| 92 |
+
"loss": 1.8281,
|
| 93 |
+
"step": 6000
|
| 94 |
+
},
|
| 95 |
+
{
|
| 96 |
+
"epoch": 0.1674049654888225,
|
| 97 |
+
"grad_norm": 9.086551666259766,
|
| 98 |
+
"learning_rate": 6.9765885443494386e-06,
|
| 99 |
+
"loss": 1.8149,
|
| 100 |
+
"step": 6500
|
| 101 |
+
},
|
| 102 |
+
{
|
| 103 |
+
"epoch": 0.18028227052642423,
|
| 104 |
+
"grad_norm": 6.515509128570557,
|
| 105 |
+
"learning_rate": 6.974785721644174e-06,
|
| 106 |
+
"loss": 1.8216,
|
| 107 |
+
"step": 7000
|
| 108 |
+
},
|
| 109 |
+
{
|
| 110 |
+
"epoch": 0.19315957556402596,
|
| 111 |
+
"grad_norm": 6.047694683074951,
|
| 112 |
+
"learning_rate": 6.97298289893891e-06,
|
| 113 |
+
"loss": 1.8194,
|
| 114 |
+
"step": 7500
|
| 115 |
+
},
|
| 116 |
+
{
|
| 117 |
+
"epoch": 0.2060368806016277,
|
| 118 |
+
"grad_norm": 7.643856525421143,
|
| 119 |
+
"learning_rate": 6.971180076233646e-06,
|
| 120 |
+
"loss": 1.8236,
|
| 121 |
+
"step": 8000
|
| 122 |
+
},
|
| 123 |
+
{
|
| 124 |
+
"epoch": 0.21891418563922943,
|
| 125 |
+
"grad_norm": 7.796109199523926,
|
| 126 |
+
"learning_rate": 6.9693808591737915e-06,
|
| 127 |
+
"loss": 1.8147,
|
| 128 |
+
"step": 8500
|
| 129 |
+
},
|
| 130 |
+
{
|
| 131 |
+
"epoch": 0.23179149067683116,
|
| 132 |
+
"grad_norm": 5.515515327453613,
|
| 133 |
+
"learning_rate": 6.967578036468528e-06,
|
| 134 |
+
"loss": 1.8178,
|
| 135 |
+
"step": 9000
|
| 136 |
+
},
|
| 137 |
+
{
|
| 138 |
+
"epoch": 0.2446687957144329,
|
| 139 |
+
"grad_norm": 8.763411521911621,
|
| 140 |
+
"learning_rate": 6.965775213763264e-06,
|
| 141 |
+
"loss": 1.8173,
|
| 142 |
+
"step": 9500
|
| 143 |
+
},
|
| 144 |
+
{
|
| 145 |
+
"epoch": 0.2575461007520346,
|
| 146 |
+
"grad_norm": 5.124142169952393,
|
| 147 |
+
"learning_rate": 6.963972391058e-06,
|
| 148 |
+
"loss": 1.8086,
|
| 149 |
+
"step": 10000
|
| 150 |
+
},
|
| 151 |
+
{
|
| 152 |
+
"epoch": 0.27042340578963636,
|
| 153 |
+
"grad_norm": 7.082516193389893,
|
| 154 |
+
"learning_rate": 6.962173173998145e-06,
|
| 155 |
+
"loss": 1.8162,
|
| 156 |
+
"step": 10500
|
| 157 |
+
},
|
| 158 |
+
{
|
| 159 |
+
"epoch": 0.28330071082723807,
|
| 160 |
+
"grad_norm": 4.873811721801758,
|
| 161 |
+
"learning_rate": 6.960370351292881e-06,
|
| 162 |
+
"loss": 1.8011,
|
| 163 |
+
"step": 11000
|
| 164 |
+
},
|
| 165 |
+
{
|
| 166 |
+
"epoch": 0.2961780158648398,
|
| 167 |
+
"grad_norm": 5.1067585945129395,
|
| 168 |
+
"learning_rate": 6.958567528587618e-06,
|
| 169 |
+
"loss": 1.8113,
|
| 170 |
+
"step": 11500
|
| 171 |
+
},
|
| 172 |
+
{
|
| 173 |
+
"epoch": 0.30905532090244153,
|
| 174 |
+
"grad_norm": 5.661747932434082,
|
| 175 |
+
"learning_rate": 6.956764705882353e-06,
|
| 176 |
+
"loss": 1.8179,
|
| 177 |
+
"step": 12000
|
| 178 |
+
},
|
| 179 |
+
{
|
| 180 |
+
"epoch": 0.3219326259400433,
|
| 181 |
+
"grad_norm": 6.289585590362549,
|
| 182 |
+
"learning_rate": 6.954965488822499e-06,
|
| 183 |
+
"loss": 1.807,
|
| 184 |
+
"step": 12500
|
| 185 |
+
},
|
| 186 |
+
{
|
| 187 |
+
"epoch": 0.334809930977645,
|
| 188 |
+
"grad_norm": 5.190219402313232,
|
| 189 |
+
"learning_rate": 6.953162666117235e-06,
|
| 190 |
+
"loss": 1.8019,
|
| 191 |
+
"step": 13000
|
| 192 |
+
},
|
| 193 |
+
{
|
| 194 |
+
"epoch": 0.3476872360152467,
|
| 195 |
+
"grad_norm": 7.0948333740234375,
|
| 196 |
+
"learning_rate": 6.951359843411971e-06,
|
| 197 |
+
"loss": 1.8114,
|
| 198 |
+
"step": 13500
|
| 199 |
+
},
|
| 200 |
+
{
|
| 201 |
+
"epoch": 0.36056454105284846,
|
| 202 |
+
"grad_norm": 8.964125633239746,
|
| 203 |
+
"learning_rate": 6.9495570207067066e-06,
|
| 204 |
+
"loss": 1.8003,
|
| 205 |
+
"step": 14000
|
| 206 |
+
},
|
| 207 |
+
{
|
| 208 |
+
"epoch": 0.37344184609045017,
|
| 209 |
+
"grad_norm": 8.000853538513184,
|
| 210 |
+
"learning_rate": 6.947757803646853e-06,
|
| 211 |
+
"loss": 1.7997,
|
| 212 |
+
"step": 14500
|
| 213 |
+
},
|
| 214 |
+
{
|
| 215 |
+
"epoch": 0.38631915112805193,
|
| 216 |
+
"grad_norm": 5.036130428314209,
|
| 217 |
+
"learning_rate": 6.945954980941588e-06,
|
| 218 |
+
"loss": 1.8046,
|
| 219 |
+
"step": 15000
|
| 220 |
+
},
|
| 221 |
+
{
|
| 222 |
+
"epoch": 0.39919645616565363,
|
| 223 |
+
"grad_norm": 5.919955730438232,
|
| 224 |
+
"learning_rate": 6.944152158236324e-06,
|
| 225 |
+
"loss": 1.8084,
|
| 226 |
+
"step": 15500
|
| 227 |
+
},
|
| 228 |
+
{
|
| 229 |
+
"epoch": 0.4120737612032554,
|
| 230 |
+
"grad_norm": 6.051347255706787,
|
| 231 |
+
"learning_rate": 6.9423493355310595e-06,
|
| 232 |
+
"loss": 1.8101,
|
| 233 |
+
"step": 16000
|
| 234 |
+
},
|
| 235 |
+
{
|
| 236 |
+
"epoch": 0.4249510662408571,
|
| 237 |
+
"grad_norm": 5.415486812591553,
|
| 238 |
+
"learning_rate": 6.940550118471206e-06,
|
| 239 |
+
"loss": 1.8002,
|
| 240 |
+
"step": 16500
|
| 241 |
+
},
|
| 242 |
+
{
|
| 243 |
+
"epoch": 0.43782837127845886,
|
| 244 |
+
"grad_norm": 5.6131744384765625,
|
| 245 |
+
"learning_rate": 6.938747295765942e-06,
|
| 246 |
+
"loss": 1.8176,
|
| 247 |
+
"step": 17000
|
| 248 |
+
},
|
| 249 |
+
{
|
| 250 |
+
"epoch": 0.45070567631606057,
|
| 251 |
+
"grad_norm": 5.342150688171387,
|
| 252 |
+
"learning_rate": 6.9369444730606775e-06,
|
| 253 |
+
"loss": 1.8064,
|
| 254 |
+
"step": 17500
|
| 255 |
+
},
|
| 256 |
+
{
|
| 257 |
+
"epoch": 0.4635829813536623,
|
| 258 |
+
"grad_norm": 8.20988655090332,
|
| 259 |
+
"learning_rate": 6.935141650355413e-06,
|
| 260 |
+
"loss": 1.8067,
|
| 261 |
+
"step": 18000
|
| 262 |
+
},
|
| 263 |
+
{
|
| 264 |
+
"epoch": 0.47646028639126403,
|
| 265 |
+
"grad_norm": 6.165362358093262,
|
| 266 |
+
"learning_rate": 6.933338827650149e-06,
|
| 267 |
+
"loss": 1.7982,
|
| 268 |
+
"step": 18500
|
| 269 |
+
},
|
| 270 |
+
{
|
| 271 |
+
"epoch": 0.4893375914288658,
|
| 272 |
+
"grad_norm": 5.723970413208008,
|
| 273 |
+
"learning_rate": 6.9315396105902955e-06,
|
| 274 |
+
"loss": 1.8078,
|
| 275 |
+
"step": 19000
|
| 276 |
+
},
|
| 277 |
+
{
|
| 278 |
+
"epoch": 0.5022148964664676,
|
| 279 |
+
"grad_norm": 5.340003967285156,
|
| 280 |
+
"learning_rate": 6.929740393530442e-06,
|
| 281 |
+
"loss": 1.8059,
|
| 282 |
+
"step": 19500
|
| 283 |
+
},
|
| 284 |
+
{
|
| 285 |
+
"epoch": 0.5150922015040692,
|
| 286 |
+
"grad_norm": 7.968153953552246,
|
| 287 |
+
"learning_rate": 6.927937570825178e-06,
|
| 288 |
+
"loss": 1.8028,
|
| 289 |
+
"step": 20000
|
| 290 |
+
},
|
| 291 |
+
{
|
| 292 |
+
"epoch": 0.527969506541671,
|
| 293 |
+
"grad_norm": 5.750583171844482,
|
| 294 |
+
"learning_rate": 6.9261347481199135e-06,
|
| 295 |
+
"loss": 1.7912,
|
| 296 |
+
"step": 20500
|
| 297 |
+
},
|
| 298 |
+
{
|
| 299 |
+
"epoch": 0.5408468115792727,
|
| 300 |
+
"grad_norm": 6.711771488189697,
|
| 301 |
+
"learning_rate": 6.924331925414649e-06,
|
| 302 |
+
"loss": 1.8043,
|
| 303 |
+
"step": 21000
|
| 304 |
+
},
|
| 305 |
+
{
|
| 306 |
+
"epoch": 0.5537241166168744,
|
| 307 |
+
"grad_norm": 6.380643367767334,
|
| 308 |
+
"learning_rate": 6.922529102709385e-06,
|
| 309 |
+
"loss": 1.7956,
|
| 310 |
+
"step": 21500
|
| 311 |
+
},
|
| 312 |
+
{
|
| 313 |
+
"epoch": 0.5666014216544761,
|
| 314 |
+
"grad_norm": 7.911167621612549,
|
| 315 |
+
"learning_rate": 6.92072628000412e-06,
|
| 316 |
+
"loss": 1.8033,
|
| 317 |
+
"step": 22000
|
| 318 |
+
},
|
| 319 |
+
{
|
| 320 |
+
"epoch": 0.5794787266920779,
|
| 321 |
+
"grad_norm": 5.734021186828613,
|
| 322 |
+
"learning_rate": 6.918923457298856e-06,
|
| 323 |
+
"loss": 1.8028,
|
| 324 |
+
"step": 22500
|
| 325 |
+
},
|
| 326 |
+
{
|
| 327 |
+
"epoch": 0.5923560317296797,
|
| 328 |
+
"grad_norm": 3.4658050537109375,
|
| 329 |
+
"learning_rate": 6.917120634593592e-06,
|
| 330 |
+
"loss": 1.8017,
|
| 331 |
+
"step": 23000
|
| 332 |
+
},
|
| 333 |
+
{
|
| 334 |
+
"epoch": 0.6052333367672813,
|
| 335 |
+
"grad_norm": 5.958466053009033,
|
| 336 |
+
"learning_rate": 6.915321417533738e-06,
|
| 337 |
+
"loss": 1.8048,
|
| 338 |
+
"step": 23500
|
| 339 |
+
},
|
| 340 |
+
{
|
| 341 |
+
"epoch": 0.6181106418048831,
|
| 342 |
+
"grad_norm": 6.3394455909729,
|
| 343 |
+
"learning_rate": 6.913518594828474e-06,
|
| 344 |
+
"loss": 1.7956,
|
| 345 |
+
"step": 24000
|
| 346 |
+
},
|
| 347 |
+
{
|
| 348 |
+
"epoch": 0.6309879468424848,
|
| 349 |
+
"grad_norm": 5.954572677612305,
|
| 350 |
+
"learning_rate": 6.91171577212321e-06,
|
| 351 |
+
"loss": 1.7944,
|
| 352 |
+
"step": 24500
|
| 353 |
+
},
|
| 354 |
+
{
|
| 355 |
+
"epoch": 0.6438652518800866,
|
| 356 |
+
"grad_norm": 7.130244731903076,
|
| 357 |
+
"learning_rate": 6.9099129494179455e-06,
|
| 358 |
+
"loss": 1.7954,
|
| 359 |
+
"step": 25000
|
| 360 |
+
},
|
| 361 |
+
{
|
| 362 |
+
"epoch": 0.6567425569176882,
|
| 363 |
+
"grad_norm": 7.7536725997924805,
|
| 364 |
+
"learning_rate": 6.908110126712681e-06,
|
| 365 |
+
"loss": 1.8015,
|
| 366 |
+
"step": 25500
|
| 367 |
+
},
|
| 368 |
+
{
|
| 369 |
+
"epoch": 0.66961986195529,
|
| 370 |
+
"grad_norm": 8.227217674255371,
|
| 371 |
+
"learning_rate": 6.906307304007418e-06,
|
| 372 |
+
"loss": 1.7917,
|
| 373 |
+
"step": 26000
|
| 374 |
+
},
|
| 375 |
+
{
|
| 376 |
+
"epoch": 0.6824971669928918,
|
| 377 |
+
"grad_norm": 5.820123195648193,
|
| 378 |
+
"learning_rate": 6.904504481302153e-06,
|
| 379 |
+
"loss": 1.7853,
|
| 380 |
+
"step": 26500
|
| 381 |
+
},
|
| 382 |
+
{
|
| 383 |
+
"epoch": 0.6953744720304934,
|
| 384 |
+
"grad_norm": 3.5955381393432617,
|
| 385 |
+
"learning_rate": 6.902705264242299e-06,
|
| 386 |
+
"loss": 1.7995,
|
| 387 |
+
"step": 27000
|
| 388 |
+
},
|
| 389 |
+
{
|
| 390 |
+
"epoch": 0.7082517770680952,
|
| 391 |
+
"grad_norm": 6.445123672485352,
|
| 392 |
+
"learning_rate": 6.900902441537035e-06,
|
| 393 |
+
"loss": 1.795,
|
| 394 |
+
"step": 27500
|
| 395 |
+
},
|
| 396 |
+
{
|
| 397 |
+
"epoch": 0.7211290821056969,
|
| 398 |
+
"grad_norm": 8.052949905395508,
|
| 399 |
+
"learning_rate": 6.89909961883177e-06,
|
| 400 |
+
"loss": 1.7967,
|
| 401 |
+
"step": 28000
|
| 402 |
+
},
|
| 403 |
+
{
|
| 404 |
+
"epoch": 0.7340063871432987,
|
| 405 |
+
"grad_norm": 4.261895179748535,
|
| 406 |
+
"learning_rate": 6.897296796126507e-06,
|
| 407 |
+
"loss": 1.789,
|
| 408 |
+
"step": 28500
|
| 409 |
+
},
|
| 410 |
+
{
|
| 411 |
+
"epoch": 0.7468836921809003,
|
| 412 |
+
"grad_norm": 5.8953633308410645,
|
| 413 |
+
"learning_rate": 6.895497579066653e-06,
|
| 414 |
+
"loss": 1.7828,
|
| 415 |
+
"step": 29000
|
| 416 |
+
},
|
| 417 |
+
{
|
| 418 |
+
"epoch": 0.7597609972185021,
|
| 419 |
+
"grad_norm": 5.249965667724609,
|
| 420 |
+
"learning_rate": 6.893694756361388e-06,
|
| 421 |
+
"loss": 1.8035,
|
| 422 |
+
"step": 29500
|
| 423 |
+
},
|
| 424 |
+
{
|
| 425 |
+
"epoch": 0.7726383022561039,
|
| 426 |
+
"grad_norm": 7.793389320373535,
|
| 427 |
+
"learning_rate": 6.891895539301535e-06,
|
| 428 |
+
"loss": 1.8054,
|
| 429 |
+
"step": 30000
|
| 430 |
+
},
|
| 431 |
+
{
|
| 432 |
+
"epoch": 0.7855156072937056,
|
| 433 |
+
"grad_norm": 5.460562229156494,
|
| 434 |
+
"learning_rate": 6.890092716596271e-06,
|
| 435 |
+
"loss": 1.7936,
|
| 436 |
+
"step": 30500
|
| 437 |
+
},
|
| 438 |
+
{
|
| 439 |
+
"epoch": 0.7983929123313073,
|
| 440 |
+
"grad_norm": 5.304771423339844,
|
| 441 |
+
"learning_rate": 6.888289893891006e-06,
|
| 442 |
+
"loss": 1.783,
|
| 443 |
+
"step": 31000
|
| 444 |
+
},
|
| 445 |
+
{
|
| 446 |
+
"epoch": 0.811270217368909,
|
| 447 |
+
"grad_norm": 5.964851379394531,
|
| 448 |
+
"learning_rate": 6.886487071185742e-06,
|
| 449 |
+
"loss": 1.7881,
|
| 450 |
+
"step": 31500
|
| 451 |
+
},
|
| 452 |
+
{
|
| 453 |
+
"epoch": 0.8241475224065108,
|
| 454 |
+
"grad_norm": 5.454218864440918,
|
| 455 |
+
"learning_rate": 6.884684248480478e-06,
|
| 456 |
+
"loss": 1.7944,
|
| 457 |
+
"step": 32000
|
| 458 |
+
},
|
| 459 |
+
{
|
| 460 |
+
"epoch": 0.8370248274441126,
|
| 461 |
+
"grad_norm": 5.067232131958008,
|
| 462 |
+
"learning_rate": 6.8828814257752135e-06,
|
| 463 |
+
"loss": 1.7939,
|
| 464 |
+
"step": 32500
|
| 465 |
+
},
|
| 466 |
+
{
|
| 467 |
+
"epoch": 0.8499021324817142,
|
| 468 |
+
"grad_norm": 6.87580680847168,
|
| 469 |
+
"learning_rate": 6.881078603069949e-06,
|
| 470 |
+
"loss": 1.7849,
|
| 471 |
+
"step": 33000
|
| 472 |
+
},
|
| 473 |
+
{
|
| 474 |
+
"epoch": 0.862779437519316,
|
| 475 |
+
"grad_norm": 7.284745693206787,
|
| 476 |
+
"learning_rate": 6.879275780364685e-06,
|
| 477 |
+
"loss": 1.7981,
|
| 478 |
+
"step": 33500
|
| 479 |
+
},
|
| 480 |
+
{
|
| 481 |
+
"epoch": 0.8756567425569177,
|
| 482 |
+
"grad_norm": 7.6209540367126465,
|
| 483 |
+
"learning_rate": 6.8774765633048315e-06,
|
| 484 |
+
"loss": 1.8011,
|
| 485 |
+
"step": 34000
|
| 486 |
+
},
|
| 487 |
+
{
|
| 488 |
+
"epoch": 0.8885340475945194,
|
| 489 |
+
"grad_norm": 6.659396648406982,
|
| 490 |
+
"learning_rate": 6.875673740599567e-06,
|
| 491 |
+
"loss": 1.8105,
|
| 492 |
+
"step": 34500
|
| 493 |
+
},
|
| 494 |
+
{
|
| 495 |
+
"epoch": 0.9014113526321211,
|
| 496 |
+
"grad_norm": 5.977092266082764,
|
| 497 |
+
"learning_rate": 6.873870917894302e-06,
|
| 498 |
+
"loss": 1.7939,
|
| 499 |
+
"step": 35000
|
| 500 |
+
},
|
| 501 |
+
{
|
| 502 |
+
"epoch": 0.9142886576697229,
|
| 503 |
+
"grad_norm": 6.564746379852295,
|
| 504 |
+
"learning_rate": 6.872068095189039e-06,
|
| 505 |
+
"loss": 1.7882,
|
| 506 |
+
"step": 35500
|
| 507 |
+
},
|
| 508 |
+
{
|
| 509 |
+
"epoch": 0.9271659627073247,
|
| 510 |
+
"grad_norm": 6.975079536437988,
|
| 511 |
+
"learning_rate": 6.870265272483775e-06,
|
| 512 |
+
"loss": 1.7942,
|
| 513 |
+
"step": 36000
|
| 514 |
+
},
|
| 515 |
+
{
|
| 516 |
+
"epoch": 0.9400432677449263,
|
| 517 |
+
"grad_norm": 8.673888206481934,
|
| 518 |
+
"learning_rate": 6.8684624497785105e-06,
|
| 519 |
+
"loss": 1.7938,
|
| 520 |
+
"step": 36500
|
| 521 |
+
},
|
| 522 |
+
{
|
| 523 |
+
"epoch": 0.9529205727825281,
|
| 524 |
+
"grad_norm": 6.115748405456543,
|
| 525 |
+
"learning_rate": 6.866663232718656e-06,
|
| 526 |
+
"loss": 1.8014,
|
| 527 |
+
"step": 37000
|
| 528 |
+
},
|
| 529 |
+
{
|
| 530 |
+
"epoch": 0.9657978778201298,
|
| 531 |
+
"grad_norm": 5.892397403717041,
|
| 532 |
+
"learning_rate": 6.864860410013392e-06,
|
| 533 |
+
"loss": 1.7955,
|
| 534 |
+
"step": 37500
|
| 535 |
+
},
|
| 536 |
+
{
|
| 537 |
+
"epoch": 0.9786751828577316,
|
| 538 |
+
"grad_norm": 3.719658374786377,
|
| 539 |
+
"learning_rate": 6.863057587308128e-06,
|
| 540 |
+
"loss": 1.773,
|
| 541 |
+
"step": 38000
|
| 542 |
+
},
|
| 543 |
+
{
|
| 544 |
+
"epoch": 0.9915524878953332,
|
| 545 |
+
"grad_norm": 6.752033233642578,
|
| 546 |
+
"learning_rate": 6.861254764602864e-06,
|
| 547 |
+
"loss": 1.7828,
|
| 548 |
+
"step": 38500
|
| 549 |
+
},
|
| 550 |
+
{
|
| 551 |
+
"epoch": 1.0,
|
| 552 |
+
"eval_accuracy": 0.2799319579535046,
|
| 553 |
+
"eval_f1_macro": 0.14142429830939493,
|
| 554 |
+
"eval_f1_micro": 0.2799319579535046,
|
| 555 |
+
"eval_loss": 1.8260160684585571,
|
| 556 |
+
"eval_runtime": 225.2986,
|
| 557 |
+
"eval_samples_per_second": 101.763,
|
| 558 |
+
"eval_steps_per_second": 12.721,
|
| 559 |
+
"step": 38828
|
| 560 |
+
},
|
| 561 |
+
{
|
| 562 |
+
"epoch": 1.0044297929329349,
|
| 563 |
+
"grad_norm": 5.212080955505371,
|
| 564 |
+
"learning_rate": 6.85945554754301e-06,
|
| 565 |
+
"loss": 1.7907,
|
| 566 |
+
"step": 39000
|
| 567 |
+
},
|
| 568 |
+
{
|
| 569 |
+
"epoch": 1.0173070979705368,
|
| 570 |
+
"grad_norm": 3.8510255813598633,
|
| 571 |
+
"learning_rate": 6.857652724837746e-06,
|
| 572 |
+
"loss": 1.8011,
|
| 573 |
+
"step": 39500
|
| 574 |
+
},
|
| 575 |
+
{
|
| 576 |
+
"epoch": 1.0301844030081384,
|
| 577 |
+
"grad_norm": 6.683523654937744,
|
| 578 |
+
"learning_rate": 6.8558499021324815e-06,
|
| 579 |
+
"loss": 1.7797,
|
| 580 |
+
"step": 40000
|
| 581 |
+
},
|
| 582 |
+
{
|
| 583 |
+
"epoch": 1.0430617080457403,
|
| 584 |
+
"grad_norm": 5.834150791168213,
|
| 585 |
+
"learning_rate": 6.854047079427217e-06,
|
| 586 |
+
"loss": 1.7872,
|
| 587 |
+
"step": 40500
|
| 588 |
+
},
|
| 589 |
+
{
|
| 590 |
+
"epoch": 1.055939013083342,
|
| 591 |
+
"grad_norm": 6.447351932525635,
|
| 592 |
+
"learning_rate": 6.852244256721953e-06,
|
| 593 |
+
"loss": 1.7769,
|
| 594 |
+
"step": 41000
|
| 595 |
+
},
|
| 596 |
+
{
|
| 597 |
+
"epoch": 1.0688163181209436,
|
| 598 |
+
"grad_norm": 7.13225793838501,
|
| 599 |
+
"learning_rate": 6.8504450396620995e-06,
|
| 600 |
+
"loss": 1.7762,
|
| 601 |
+
"step": 41500
|
| 602 |
+
},
|
| 603 |
+
{
|
| 604 |
+
"epoch": 1.0816936231585454,
|
| 605 |
+
"grad_norm": 4.302614212036133,
|
| 606 |
+
"learning_rate": 6.848642216956835e-06,
|
| 607 |
+
"loss": 1.7885,
|
| 608 |
+
"step": 42000
|
| 609 |
+
},
|
| 610 |
+
{
|
| 611 |
+
"epoch": 1.094570928196147,
|
| 612 |
+
"grad_norm": 4.33312463760376,
|
| 613 |
+
"learning_rate": 6.84683939425157e-06,
|
| 614 |
+
"loss": 1.7955,
|
| 615 |
+
"step": 42500
|
| 616 |
+
},
|
| 617 |
+
{
|
| 618 |
+
"epoch": 1.1074482332337487,
|
| 619 |
+
"grad_norm": 4.740920066833496,
|
| 620 |
+
"learning_rate": 6.845036571546307e-06,
|
| 621 |
+
"loss": 1.7874,
|
| 622 |
+
"step": 43000
|
| 623 |
+
},
|
| 624 |
+
{
|
| 625 |
+
"epoch": 1.1203255382713506,
|
| 626 |
+
"grad_norm": 3.6999611854553223,
|
| 627 |
+
"learning_rate": 6.843233748841043e-06,
|
| 628 |
+
"loss": 1.7879,
|
| 629 |
+
"step": 43500
|
| 630 |
+
},
|
| 631 |
+
{
|
| 632 |
+
"epoch": 1.1332028433089523,
|
| 633 |
+
"grad_norm": 4.698836803436279,
|
| 634 |
+
"learning_rate": 6.8414309261357785e-06,
|
| 635 |
+
"loss": 1.7902,
|
| 636 |
+
"step": 44000
|
| 637 |
+
},
|
| 638 |
+
{
|
| 639 |
+
"epoch": 1.1460801483465541,
|
| 640 |
+
"grad_norm": 4.617445945739746,
|
| 641 |
+
"learning_rate": 6.839631709075924e-06,
|
| 642 |
+
"loss": 1.7941,
|
| 643 |
+
"step": 44500
|
| 644 |
+
},
|
| 645 |
+
{
|
| 646 |
+
"epoch": 1.1589574533841558,
|
| 647 |
+
"grad_norm": 4.726156234741211,
|
| 648 |
+
"learning_rate": 6.83782888637066e-06,
|
| 649 |
+
"loss": 1.8009,
|
| 650 |
+
"step": 45000
|
| 651 |
+
},
|
| 652 |
+
{
|
| 653 |
+
"epoch": 1.1718347584217574,
|
| 654 |
+
"grad_norm": 7.927763938903809,
|
| 655 |
+
"learning_rate": 6.8360260636653965e-06,
|
| 656 |
+
"loss": 1.7896,
|
| 657 |
+
"step": 45500
|
| 658 |
+
},
|
| 659 |
+
{
|
| 660 |
+
"epoch": 1.1847120634593593,
|
| 661 |
+
"grad_norm": 3.4403719902038574,
|
| 662 |
+
"learning_rate": 6.834223240960132e-06,
|
| 663 |
+
"loss": 1.796,
|
| 664 |
+
"step": 46000
|
| 665 |
+
},
|
| 666 |
+
{
|
| 667 |
+
"epoch": 1.197589368496961,
|
| 668 |
+
"grad_norm": 4.452911376953125,
|
| 669 |
+
"learning_rate": 6.832420418254867e-06,
|
| 670 |
+
"loss": 1.7789,
|
| 671 |
+
"step": 46500
|
| 672 |
+
},
|
| 673 |
+
{
|
| 674 |
+
"epoch": 1.2104666735345626,
|
| 675 |
+
"grad_norm": 6.698709487915039,
|
| 676 |
+
"learning_rate": 6.830621201195014e-06,
|
| 677 |
+
"loss": 1.7847,
|
| 678 |
+
"step": 47000
|
| 679 |
+
},
|
| 680 |
+
{
|
| 681 |
+
"epoch": 1.2233439785721645,
|
| 682 |
+
"grad_norm": 6.807077884674072,
|
| 683 |
+
"learning_rate": 6.8288183784897495e-06,
|
| 684 |
+
"loss": 1.7809,
|
| 685 |
+
"step": 47500
|
| 686 |
+
},
|
| 687 |
+
{
|
| 688 |
+
"epoch": 1.2362212836097661,
|
| 689 |
+
"grad_norm": 5.895312309265137,
|
| 690 |
+
"learning_rate": 6.827015555784485e-06,
|
| 691 |
+
"loss": 1.7701,
|
| 692 |
+
"step": 48000
|
| 693 |
+
},
|
| 694 |
+
{
|
| 695 |
+
"epoch": 1.2490985886473678,
|
| 696 |
+
"grad_norm": 4.329184055328369,
|
| 697 |
+
"learning_rate": 6.825212733079221e-06,
|
| 698 |
+
"loss": 1.7972,
|
| 699 |
+
"step": 48500
|
| 700 |
+
},
|
| 701 |
+
{
|
| 702 |
+
"epoch": 1.2619758936849697,
|
| 703 |
+
"grad_norm": 7.711607456207275,
|
| 704 |
+
"learning_rate": 6.8234135160193675e-06,
|
| 705 |
+
"loss": 1.7655,
|
| 706 |
+
"step": 49000
|
| 707 |
+
},
|
| 708 |
+
{
|
| 709 |
+
"epoch": 1.2748531987225713,
|
| 710 |
+
"grad_norm": 7.793875694274902,
|
| 711 |
+
"learning_rate": 6.8216106933141025e-06,
|
| 712 |
+
"loss": 1.7942,
|
| 713 |
+
"step": 49500
|
| 714 |
+
},
|
| 715 |
+
{
|
| 716 |
+
"epoch": 1.287730503760173,
|
| 717 |
+
"grad_norm": 5.680514812469482,
|
| 718 |
+
"learning_rate": 6.819807870608838e-06,
|
| 719 |
+
"loss": 1.7883,
|
| 720 |
+
"step": 50000
|
| 721 |
+
},
|
| 722 |
+
{
|
| 723 |
+
"epoch": 1.3006078087977748,
|
| 724 |
+
"grad_norm": 5.679476261138916,
|
| 725 |
+
"learning_rate": 6.818005047903575e-06,
|
| 726 |
+
"loss": 1.7771,
|
| 727 |
+
"step": 50500
|
| 728 |
+
},
|
| 729 |
+
{
|
| 730 |
+
"epoch": 1.3134851138353765,
|
| 731 |
+
"grad_norm": 7.666828632354736,
|
| 732 |
+
"learning_rate": 6.816202225198311e-06,
|
| 733 |
+
"loss": 1.7921,
|
| 734 |
+
"step": 51000
|
| 735 |
+
},
|
| 736 |
+
{
|
| 737 |
+
"epoch": 1.3263624188729783,
|
| 738 |
+
"grad_norm": 6.097265243530273,
|
| 739 |
+
"learning_rate": 6.814403008138456e-06,
|
| 740 |
+
"loss": 1.7928,
|
| 741 |
+
"step": 51500
|
| 742 |
+
},
|
| 743 |
+
{
|
| 744 |
+
"epoch": 1.33923972391058,
|
| 745 |
+
"grad_norm": 4.842618942260742,
|
| 746 |
+
"learning_rate": 6.812600185433192e-06,
|
| 747 |
+
"loss": 1.783,
|
| 748 |
+
"step": 52000
|
| 749 |
+
},
|
| 750 |
+
{
|
| 751 |
+
"epoch": 1.3521170289481816,
|
| 752 |
+
"grad_norm": 3.718636989593506,
|
| 753 |
+
"learning_rate": 6.810797362727928e-06,
|
| 754 |
+
"loss": 1.7889,
|
| 755 |
+
"step": 52500
|
| 756 |
+
},
|
| 757 |
+
{
|
| 758 |
+
"epoch": 1.3649943339857835,
|
| 759 |
+
"grad_norm": 5.283341884613037,
|
| 760 |
+
"learning_rate": 6.8089945400226645e-06,
|
| 761 |
+
"loss": 1.7918,
|
| 762 |
+
"step": 53000
|
| 763 |
+
},
|
| 764 |
+
{
|
| 765 |
+
"epoch": 1.3778716390233852,
|
| 766 |
+
"grad_norm": 5.642731189727783,
|
| 767 |
+
"learning_rate": 6.80719532296281e-06,
|
| 768 |
+
"loss": 1.7918,
|
| 769 |
+
"step": 53500
|
| 770 |
+
},
|
| 771 |
+
{
|
| 772 |
+
"epoch": 1.390748944060987,
|
| 773 |
+
"grad_norm": 8.50252914428711,
|
| 774 |
+
"learning_rate": 6.805392500257546e-06,
|
| 775 |
+
"loss": 1.7879,
|
| 776 |
+
"step": 54000
|
| 777 |
+
},
|
| 778 |
+
{
|
| 779 |
+
"epoch": 1.4036262490985887,
|
| 780 |
+
"grad_norm": 7.05935525894165,
|
| 781 |
+
"learning_rate": 6.803589677552282e-06,
|
| 782 |
+
"loss": 1.7857,
|
| 783 |
+
"step": 54500
|
| 784 |
+
},
|
| 785 |
+
{
|
| 786 |
+
"epoch": 1.4165035541361903,
|
| 787 |
+
"grad_norm": 4.355367183685303,
|
| 788 |
+
"learning_rate": 6.8017868548470175e-06,
|
| 789 |
+
"loss": 1.7781,
|
| 790 |
+
"step": 55000
|
| 791 |
+
},
|
| 792 |
+
{
|
| 793 |
+
"epoch": 1.4293808591737922,
|
| 794 |
+
"grad_norm": 3.4202096462249756,
|
| 795 |
+
"learning_rate": 6.799984032141753e-06,
|
| 796 |
+
"loss": 1.7979,
|
| 797 |
+
"step": 55500
|
| 798 |
+
},
|
| 799 |
+
{
|
| 800 |
+
"epoch": 1.4422581642113939,
|
| 801 |
+
"grad_norm": 4.578690052032471,
|
| 802 |
+
"learning_rate": 6.7981848150819e-06,
|
| 803 |
+
"loss": 1.7862,
|
| 804 |
+
"step": 56000
|
| 805 |
+
},
|
| 806 |
+
{
|
| 807 |
+
"epoch": 1.4551354692489955,
|
| 808 |
+
"grad_norm": 6.677925109863281,
|
| 809 |
+
"learning_rate": 6.7963819923766355e-06,
|
| 810 |
+
"loss": 1.7747,
|
| 811 |
+
"step": 56500
|
| 812 |
+
},
|
| 813 |
+
{
|
| 814 |
+
"epoch": 1.4680127742865974,
|
| 815 |
+
"grad_norm": 8.975627899169922,
|
| 816 |
+
"learning_rate": 6.7945791696713704e-06,
|
| 817 |
+
"loss": 1.8019,
|
| 818 |
+
"step": 57000
|
| 819 |
+
},
|
| 820 |
+
{
|
| 821 |
+
"epoch": 1.480890079324199,
|
| 822 |
+
"grad_norm": 4.4174275398254395,
|
| 823 |
+
"learning_rate": 6.792776346966106e-06,
|
| 824 |
+
"loss": 1.8029,
|
| 825 |
+
"step": 57500
|
| 826 |
+
},
|
| 827 |
+
{
|
| 828 |
+
"epoch": 1.4937673843618007,
|
| 829 |
+
"grad_norm": 5.21982479095459,
|
| 830 |
+
"learning_rate": 6.790973524260843e-06,
|
| 831 |
+
"loss": 1.788,
|
| 832 |
+
"step": 58000
|
| 833 |
+
},
|
| 834 |
+
{
|
| 835 |
+
"epoch": 1.5066446893994025,
|
| 836 |
+
"grad_norm": 13.477604866027832,
|
| 837 |
+
"learning_rate": 6.789170701555579e-06,
|
| 838 |
+
"loss": 1.781,
|
| 839 |
+
"step": 58500
|
| 840 |
+
},
|
| 841 |
+
{
|
| 842 |
+
"epoch": 1.5195219944370042,
|
| 843 |
+
"grad_norm": 6.470083236694336,
|
| 844 |
+
"learning_rate": 6.787371484495724e-06,
|
| 845 |
+
"loss": 1.7786,
|
| 846 |
+
"step": 59000
|
| 847 |
+
},
|
| 848 |
+
{
|
| 849 |
+
"epoch": 1.5323992994746058,
|
| 850 |
+
"grad_norm": 6.517702102661133,
|
| 851 |
+
"learning_rate": 6.78556866179046e-06,
|
| 852 |
+
"loss": 1.7833,
|
| 853 |
+
"step": 59500
|
| 854 |
+
},
|
| 855 |
+
{
|
| 856 |
+
"epoch": 1.5452766045122077,
|
| 857 |
+
"grad_norm": 5.956171035766602,
|
| 858 |
+
"learning_rate": 6.783765839085196e-06,
|
| 859 |
+
"loss": 1.7729,
|
| 860 |
+
"step": 60000
|
| 861 |
+
},
|
| 862 |
+
{
|
| 863 |
+
"epoch": 1.5581539095498094,
|
| 864 |
+
"grad_norm": 6.486398696899414,
|
| 865 |
+
"learning_rate": 6.7819630163799325e-06,
|
| 866 |
+
"loss": 1.7973,
|
| 867 |
+
"step": 60500
|
| 868 |
+
},
|
| 869 |
+
{
|
| 870 |
+
"epoch": 1.571031214587411,
|
| 871 |
+
"grad_norm": 4.920289993286133,
|
| 872 |
+
"learning_rate": 6.7801601936746675e-06,
|
| 873 |
+
"loss": 1.781,
|
| 874 |
+
"step": 61000
|
| 875 |
+
},
|
| 876 |
+
{
|
| 877 |
+
"epoch": 1.583908519625013,
|
| 878 |
+
"grad_norm": 5.362635612487793,
|
| 879 |
+
"learning_rate": 6.778357370969403e-06,
|
| 880 |
+
"loss": 1.7889,
|
| 881 |
+
"step": 61500
|
| 882 |
+
},
|
| 883 |
+
{
|
| 884 |
+
"epoch": 1.5967858246626148,
|
| 885 |
+
"grad_norm": 6.79608678817749,
|
| 886 |
+
"learning_rate": 6.776554548264139e-06,
|
| 887 |
+
"loss": 1.7926,
|
| 888 |
+
"step": 62000
|
| 889 |
+
},
|
| 890 |
+
{
|
| 891 |
+
"epoch": 1.6096631297002162,
|
| 892 |
+
"grad_norm": 4.749091625213623,
|
| 893 |
+
"learning_rate": 6.774751725558875e-06,
|
| 894 |
+
"loss": 1.7783,
|
| 895 |
+
"step": 62500
|
| 896 |
+
},
|
| 897 |
+
{
|
| 898 |
+
"epoch": 1.622540434737818,
|
| 899 |
+
"grad_norm": 7.530384540557861,
|
| 900 |
+
"learning_rate": 6.772956114144432e-06,
|
| 901 |
+
"loss": 1.7833,
|
| 902 |
+
"step": 63000
|
| 903 |
+
},
|
| 904 |
+
{
|
| 905 |
+
"epoch": 1.63541773977542,
|
| 906 |
+
"grad_norm": 6.624202251434326,
|
| 907 |
+
"learning_rate": 6.771153291439168e-06,
|
| 908 |
+
"loss": 1.7739,
|
| 909 |
+
"step": 63500
|
| 910 |
+
},
|
| 911 |
+
{
|
| 912 |
+
"epoch": 1.6482950448130216,
|
| 913 |
+
"grad_norm": 3.6536526679992676,
|
| 914 |
+
"learning_rate": 6.769350468733903e-06,
|
| 915 |
+
"loss": 1.7856,
|
| 916 |
+
"step": 64000
|
| 917 |
+
},
|
| 918 |
+
{
|
| 919 |
+
"epoch": 1.6611723498506232,
|
| 920 |
+
"grad_norm": 6.610978603363037,
|
| 921 |
+
"learning_rate": 6.7675476460286384e-06,
|
| 922 |
+
"loss": 1.7854,
|
| 923 |
+
"step": 64500
|
| 924 |
+
},
|
| 925 |
+
{
|
| 926 |
+
"epoch": 1.674049654888225,
|
| 927 |
+
"grad_norm": 5.544578552246094,
|
| 928 |
+
"learning_rate": 6.765744823323375e-06,
|
| 929 |
+
"loss": 1.774,
|
| 930 |
+
"step": 65000
|
| 931 |
+
},
|
| 932 |
+
{
|
| 933 |
+
"epoch": 1.6869269599258268,
|
| 934 |
+
"grad_norm": 7.2254133224487305,
|
| 935 |
+
"learning_rate": 6.763945606263521e-06,
|
| 936 |
+
"loss": 1.7946,
|
| 937 |
+
"step": 65500
|
| 938 |
+
},
|
| 939 |
+
{
|
| 940 |
+
"epoch": 1.6998042649634284,
|
| 941 |
+
"grad_norm": 4.5243730545043945,
|
| 942 |
+
"learning_rate": 6.7621427835582564e-06,
|
| 943 |
+
"loss": 1.8038,
|
| 944 |
+
"step": 66000
|
| 945 |
+
},
|
| 946 |
+
{
|
| 947 |
+
"epoch": 1.7126815700010303,
|
| 948 |
+
"grad_norm": 5.477443218231201,
|
| 949 |
+
"learning_rate": 6.760343566498404e-06,
|
| 950 |
+
"loss": 1.8058,
|
| 951 |
+
"step": 66500
|
| 952 |
+
},
|
| 953 |
+
{
|
| 954 |
+
"epoch": 1.725558875038632,
|
| 955 |
+
"grad_norm": 5.0484771728515625,
|
| 956 |
+
"learning_rate": 6.758540743793139e-06,
|
| 957 |
+
"loss": 1.7836,
|
| 958 |
+
"step": 67000
|
| 959 |
+
},
|
| 960 |
+
{
|
| 961 |
+
"epoch": 1.7384361800762336,
|
| 962 |
+
"grad_norm": 3.814950466156006,
|
| 963 |
+
"learning_rate": 6.7567379210878744e-06,
|
| 964 |
+
"loss": 1.7774,
|
| 965 |
+
"step": 67500
|
| 966 |
+
},
|
| 967 |
+
{
|
| 968 |
+
"epoch": 1.7513134851138354,
|
| 969 |
+
"grad_norm": 4.360231399536133,
|
| 970 |
+
"learning_rate": 6.75493509838261e-06,
|
| 971 |
+
"loss": 1.7959,
|
| 972 |
+
"step": 68000
|
| 973 |
+
},
|
| 974 |
+
{
|
| 975 |
+
"epoch": 1.764190790151437,
|
| 976 |
+
"grad_norm": 4.7067646980285645,
|
| 977 |
+
"learning_rate": 6.753132275677346e-06,
|
| 978 |
+
"loss": 1.7942,
|
| 979 |
+
"step": 68500
|
| 980 |
+
},
|
| 981 |
+
{
|
| 982 |
+
"epoch": 1.7770680951890387,
|
| 983 |
+
"grad_norm": 4.299397945404053,
|
| 984 |
+
"learning_rate": 6.751329452972082e-06,
|
| 985 |
+
"loss": 1.7847,
|
| 986 |
+
"step": 69000
|
| 987 |
+
},
|
| 988 |
+
{
|
| 989 |
+
"epoch": 1.7899454002266406,
|
| 990 |
+
"grad_norm": 4.759008407592773,
|
| 991 |
+
"learning_rate": 6.749526630266818e-06,
|
| 992 |
+
"loss": 1.7788,
|
| 993 |
+
"step": 69500
|
| 994 |
+
},
|
| 995 |
+
{
|
| 996 |
+
"epoch": 1.8028227052642423,
|
| 997 |
+
"grad_norm": 5.2251787185668945,
|
| 998 |
+
"learning_rate": 6.7477238075615535e-06,
|
| 999 |
+
"loss": 1.7998,
|
| 1000 |
+
"step": 70000
|
| 1001 |
+
},
|
| 1002 |
+
{
|
| 1003 |
+
"epoch": 1.815700010301844,
|
| 1004 |
+
"grad_norm": 6.724719524383545,
|
| 1005 |
+
"learning_rate": 6.745920984856289e-06,
|
| 1006 |
+
"loss": 1.7823,
|
| 1007 |
+
"step": 70500
|
| 1008 |
+
},
|
| 1009 |
+
{
|
| 1010 |
+
"epoch": 1.8285773153394458,
|
| 1011 |
+
"grad_norm": 4.814666271209717,
|
| 1012 |
+
"learning_rate": 6.744121767796436e-06,
|
| 1013 |
+
"loss": 1.7828,
|
| 1014 |
+
"step": 71000
|
| 1015 |
+
},
|
| 1016 |
+
{
|
| 1017 |
+
"epoch": 1.8414546203770477,
|
| 1018 |
+
"grad_norm": 4.108809947967529,
|
| 1019 |
+
"learning_rate": 6.742318945091171e-06,
|
| 1020 |
+
"loss": 1.7811,
|
| 1021 |
+
"step": 71500
|
| 1022 |
+
},
|
| 1023 |
+
{
|
| 1024 |
+
"epoch": 1.854331925414649,
|
| 1025 |
+
"grad_norm": 5.329784393310547,
|
| 1026 |
+
"learning_rate": 6.7405161223859064e-06,
|
| 1027 |
+
"loss": 1.7808,
|
| 1028 |
+
"step": 72000
|
| 1029 |
+
},
|
| 1030 |
+
{
|
| 1031 |
+
"epoch": 1.867209230452251,
|
| 1032 |
+
"grad_norm": 5.064823150634766,
|
| 1033 |
+
"learning_rate": 6.738713299680643e-06,
|
| 1034 |
+
"loss": 1.779,
|
| 1035 |
+
"step": 72500
|
| 1036 |
+
},
|
| 1037 |
+
{
|
| 1038 |
+
"epoch": 1.8800865354898528,
|
| 1039 |
+
"grad_norm": 4.863822937011719,
|
| 1040 |
+
"learning_rate": 6.736914082620789e-06,
|
| 1041 |
+
"loss": 1.7885,
|
| 1042 |
+
"step": 73000
|
| 1043 |
+
},
|
| 1044 |
+
{
|
| 1045 |
+
"epoch": 1.8929638405274543,
|
| 1046 |
+
"grad_norm": 8.565622329711914,
|
| 1047 |
+
"learning_rate": 6.7351112599155244e-06,
|
| 1048 |
+
"loss": 1.7841,
|
| 1049 |
+
"step": 73500
|
| 1050 |
+
},
|
| 1051 |
+
{
|
| 1052 |
+
"epoch": 1.9058411455650561,
|
| 1053 |
+
"grad_norm": 5.2519145011901855,
|
| 1054 |
+
"learning_rate": 6.73330843721026e-06,
|
| 1055 |
+
"loss": 1.7882,
|
| 1056 |
+
"step": 74000
|
| 1057 |
+
},
|
| 1058 |
+
{
|
| 1059 |
+
"epoch": 1.918718450602658,
|
| 1060 |
+
"grad_norm": 8.66575813293457,
|
| 1061 |
+
"learning_rate": 6.731505614504996e-06,
|
| 1062 |
+
"loss": 1.771,
|
| 1063 |
+
"step": 74500
|
| 1064 |
+
},
|
| 1065 |
+
{
|
| 1066 |
+
"epoch": 1.9315957556402596,
|
| 1067 |
+
"grad_norm": 9.521093368530273,
|
| 1068 |
+
"learning_rate": 6.729702791799733e-06,
|
| 1069 |
+
"loss": 1.7866,
|
| 1070 |
+
"step": 75000
|
| 1071 |
+
},
|
| 1072 |
+
{
|
| 1073 |
+
"epoch": 1.9444730606778613,
|
| 1074 |
+
"grad_norm": 4.49517297744751,
|
| 1075 |
+
"learning_rate": 6.727899969094468e-06,
|
| 1076 |
+
"loss": 1.7867,
|
| 1077 |
+
"step": 75500
|
| 1078 |
+
},
|
| 1079 |
+
{
|
| 1080 |
+
"epoch": 1.9573503657154632,
|
| 1081 |
+
"grad_norm": 7.0434699058532715,
|
| 1082 |
+
"learning_rate": 6.726100752034614e-06,
|
| 1083 |
+
"loss": 1.7806,
|
| 1084 |
+
"step": 76000
|
| 1085 |
+
},
|
| 1086 |
+
{
|
| 1087 |
+
"epoch": 1.9702276707530648,
|
| 1088 |
+
"grad_norm": 4.631639003753662,
|
| 1089 |
+
"learning_rate": 6.72429792932935e-06,
|
| 1090 |
+
"loss": 1.7965,
|
| 1091 |
+
"step": 76500
|
| 1092 |
+
},
|
| 1093 |
+
{
|
| 1094 |
+
"epoch": 1.9831049757906665,
|
| 1095 |
+
"grad_norm": 6.64503288269043,
|
| 1096 |
+
"learning_rate": 6.722495106624085e-06,
|
| 1097 |
+
"loss": 1.7866,
|
| 1098 |
+
"step": 77000
|
| 1099 |
+
},
|
| 1100 |
+
{
|
| 1101 |
+
"epoch": 1.9959822808282683,
|
| 1102 |
+
"grad_norm": 6.729311943054199,
|
| 1103 |
+
"learning_rate": 6.7206922839188215e-06,
|
| 1104 |
+
"loss": 1.7776,
|
| 1105 |
+
"step": 77500
|
| 1106 |
+
},
|
| 1107 |
+
{
|
| 1108 |
+
"epoch": 2.0,
|
| 1109 |
+
"eval_accuracy": 0.2799319579535046,
|
| 1110 |
+
"eval_f1_macro": 0.14207620906161436,
|
| 1111 |
+
"eval_f1_micro": 0.2799319579535046,
|
| 1112 |
+
"eval_loss": 1.7851710319519043,
|
| 1113 |
+
"eval_runtime": 223.7334,
|
| 1114 |
+
"eval_samples_per_second": 102.475,
|
| 1115 |
+
"eval_steps_per_second": 12.81,
|
| 1116 |
+
"step": 77656
|
| 1117 |
+
},
|
| 1118 |
+
{
|
| 1119 |
+
"epoch": 2.0088595858658698,
|
| 1120 |
+
"grad_norm": 5.343408584594727,
|
| 1121 |
+
"learning_rate": 6.718893066858968e-06,
|
| 1122 |
+
"loss": 1.7612,
|
| 1123 |
+
"step": 78000
|
| 1124 |
+
},
|
| 1125 |
+
{
|
| 1126 |
+
"epoch": 2.0217368909034716,
|
| 1127 |
+
"grad_norm": 7.235130786895752,
|
| 1128 |
+
"learning_rate": 6.717090244153703e-06,
|
| 1129 |
+
"loss": 1.7718,
|
| 1130 |
+
"step": 78500
|
| 1131 |
+
},
|
| 1132 |
+
{
|
| 1133 |
+
"epoch": 2.0346141959410735,
|
| 1134 |
+
"grad_norm": 3.554536819458008,
|
| 1135 |
+
"learning_rate": 6.715287421448439e-06,
|
| 1136 |
+
"loss": 1.7887,
|
| 1137 |
+
"step": 79000
|
| 1138 |
+
},
|
| 1139 |
+
{
|
| 1140 |
+
"epoch": 2.0474915009786754,
|
| 1141 |
+
"grad_norm": 6.864304542541504,
|
| 1142 |
+
"learning_rate": 6.7134845987431744e-06,
|
| 1143 |
+
"loss": 1.7795,
|
| 1144 |
+
"step": 79500
|
| 1145 |
+
},
|
| 1146 |
+
{
|
| 1147 |
+
"epoch": 2.060368806016277,
|
| 1148 |
+
"grad_norm": 4.944849967956543,
|
| 1149 |
+
"learning_rate": 6.711681776037911e-06,
|
| 1150 |
+
"loss": 1.7826,
|
| 1151 |
+
"step": 80000
|
| 1152 |
+
},
|
| 1153 |
+
{
|
| 1154 |
+
"epoch": 2.0732461110538787,
|
| 1155 |
+
"grad_norm": 4.683295726776123,
|
| 1156 |
+
"learning_rate": 6.709878953332647e-06,
|
| 1157 |
+
"loss": 1.7711,
|
| 1158 |
+
"step": 80500
|
| 1159 |
+
},
|
| 1160 |
+
{
|
| 1161 |
+
"epoch": 2.0861234160914806,
|
| 1162 |
+
"grad_norm": 5.293025493621826,
|
| 1163 |
+
"learning_rate": 6.708076130627383e-06,
|
| 1164 |
+
"loss": 1.7863,
|
| 1165 |
+
"step": 81000
|
| 1166 |
+
},
|
| 1167 |
+
{
|
| 1168 |
+
"epoch": 2.099000721129082,
|
| 1169 |
+
"grad_norm": 3.2971994876861572,
|
| 1170 |
+
"learning_rate": 6.706273307922118e-06,
|
| 1171 |
+
"loss": 1.7759,
|
| 1172 |
+
"step": 81500
|
| 1173 |
+
},
|
| 1174 |
+
{
|
| 1175 |
+
"epoch": 2.111878026166684,
|
| 1176 |
+
"grad_norm": 4.406072616577148,
|
| 1177 |
+
"learning_rate": 6.704474090862264e-06,
|
| 1178 |
+
"loss": 1.7828,
|
| 1179 |
+
"step": 82000
|
| 1180 |
+
},
|
| 1181 |
+
{
|
| 1182 |
+
"epoch": 2.1247553312042857,
|
| 1183 |
+
"grad_norm": 5.096445083618164,
|
| 1184 |
+
"learning_rate": 6.7026748738024104e-06,
|
| 1185 |
+
"loss": 1.7764,
|
| 1186 |
+
"step": 82500
|
| 1187 |
+
},
|
| 1188 |
+
{
|
| 1189 |
+
"epoch": 2.137632636241887,
|
| 1190 |
+
"grad_norm": 5.056951522827148,
|
| 1191 |
+
"learning_rate": 6.700872051097146e-06,
|
| 1192 |
+
"loss": 1.7884,
|
| 1193 |
+
"step": 83000
|
| 1194 |
+
},
|
| 1195 |
+
{
|
| 1196 |
+
"epoch": 2.150509941279489,
|
| 1197 |
+
"grad_norm": 5.062712669372559,
|
| 1198 |
+
"learning_rate": 6.699069228391882e-06,
|
| 1199 |
+
"loss": 1.7772,
|
| 1200 |
+
"step": 83500
|
| 1201 |
+
},
|
| 1202 |
+
{
|
| 1203 |
+
"epoch": 2.163387246317091,
|
| 1204 |
+
"grad_norm": 5.221599578857422,
|
| 1205 |
+
"learning_rate": 6.697266405686618e-06,
|
| 1206 |
+
"loss": 1.7845,
|
| 1207 |
+
"step": 84000
|
| 1208 |
+
},
|
| 1209 |
+
{
|
| 1210 |
+
"epoch": 2.1762645513546923,
|
| 1211 |
+
"grad_norm": 6.651246547698975,
|
| 1212 |
+
"learning_rate": 6.695463582981354e-06,
|
| 1213 |
+
"loss": 1.7873,
|
| 1214 |
+
"step": 84500
|
| 1215 |
+
},
|
| 1216 |
+
{
|
| 1217 |
+
"epoch": 2.189141856392294,
|
| 1218 |
+
"grad_norm": 3.7498083114624023,
|
| 1219 |
+
"learning_rate": 6.6936607602760895e-06,
|
| 1220 |
+
"loss": 1.7729,
|
| 1221 |
+
"step": 85000
|
| 1222 |
+
},
|
| 1223 |
+
{
|
| 1224 |
+
"epoch": 2.202019161429896,
|
| 1225 |
+
"grad_norm": 3.9556636810302734,
|
| 1226 |
+
"learning_rate": 6.691857937570825e-06,
|
| 1227 |
+
"loss": 1.7914,
|
| 1228 |
+
"step": 85500
|
| 1229 |
+
},
|
| 1230 |
+
{
|
| 1231 |
+
"epoch": 2.2148964664674975,
|
| 1232 |
+
"grad_norm": 4.400400638580322,
|
| 1233 |
+
"learning_rate": 6.690058720510971e-06,
|
| 1234 |
+
"loss": 1.7878,
|
| 1235 |
+
"step": 86000
|
| 1236 |
+
},
|
| 1237 |
+
{
|
| 1238 |
+
"epoch": 2.2277737715050994,
|
| 1239 |
+
"grad_norm": 5.495286464691162,
|
| 1240 |
+
"learning_rate": 6.688255897805707e-06,
|
| 1241 |
+
"loss": 1.7894,
|
| 1242 |
+
"step": 86500
|
| 1243 |
+
},
|
| 1244 |
+
{
|
| 1245 |
+
"epoch": 2.2406510765427012,
|
| 1246 |
+
"grad_norm": 5.062817573547363,
|
| 1247 |
+
"learning_rate": 6.686453075100443e-06,
|
| 1248 |
+
"loss": 1.78,
|
| 1249 |
+
"step": 87000
|
| 1250 |
+
},
|
| 1251 |
+
{
|
| 1252 |
+
"epoch": 2.2535283815803027,
|
| 1253 |
+
"grad_norm": 7.037624835968018,
|
| 1254 |
+
"learning_rate": 6.684650252395179e-06,
|
| 1255 |
+
"loss": 1.7799,
|
| 1256 |
+
"step": 87500
|
| 1257 |
+
},
|
| 1258 |
+
{
|
| 1259 |
+
"epoch": 2.2664056866179045,
|
| 1260 |
+
"grad_norm": 5.7241291999816895,
|
| 1261 |
+
"learning_rate": 6.682851035335325e-06,
|
| 1262 |
+
"loss": 1.7864,
|
| 1263 |
+
"step": 88000
|
| 1264 |
+
},
|
| 1265 |
+
{
|
| 1266 |
+
"epoch": 2.2792829916555064,
|
| 1267 |
+
"grad_norm": 6.502715587615967,
|
| 1268 |
+
"learning_rate": 6.68104821263006e-06,
|
| 1269 |
+
"loss": 1.7818,
|
| 1270 |
+
"step": 88500
|
| 1271 |
+
},
|
| 1272 |
+
{
|
| 1273 |
+
"epoch": 2.2921602966931083,
|
| 1274 |
+
"grad_norm": 9.763015747070312,
|
| 1275 |
+
"learning_rate": 6.679245389924796e-06,
|
| 1276 |
+
"loss": 1.7673,
|
| 1277 |
+
"step": 89000
|
| 1278 |
+
},
|
| 1279 |
+
{
|
| 1280 |
+
"epoch": 2.3050376017307097,
|
| 1281 |
+
"grad_norm": 6.935070991516113,
|
| 1282 |
+
"learning_rate": 6.677442567219532e-06,
|
| 1283 |
+
"loss": 1.7659,
|
| 1284 |
+
"step": 89500
|
| 1285 |
+
},
|
| 1286 |
+
{
|
| 1287 |
+
"epoch": 2.3179149067683116,
|
| 1288 |
+
"grad_norm": 9.335469245910645,
|
| 1289 |
+
"learning_rate": 6.675639744514268e-06,
|
| 1290 |
+
"loss": 1.7795,
|
| 1291 |
+
"step": 90000
|
| 1292 |
+
},
|
| 1293 |
+
{
|
| 1294 |
+
"epoch": 2.3307922118059134,
|
| 1295 |
+
"grad_norm": 7.891449928283691,
|
| 1296 |
+
"learning_rate": 6.673836921809004e-06,
|
| 1297 |
+
"loss": 1.778,
|
| 1298 |
+
"step": 90500
|
| 1299 |
+
},
|
| 1300 |
+
{
|
| 1301 |
+
"epoch": 2.343669516843515,
|
| 1302 |
+
"grad_norm": 4.85399055480957,
|
| 1303 |
+
"learning_rate": 6.67203770474915e-06,
|
| 1304 |
+
"loss": 1.782,
|
| 1305 |
+
"step": 91000
|
| 1306 |
+
},
|
| 1307 |
+
{
|
| 1308 |
+
"epoch": 2.3565468218811167,
|
| 1309 |
+
"grad_norm": 6.049405574798584,
|
| 1310 |
+
"learning_rate": 6.670234882043885e-06,
|
| 1311 |
+
"loss": 1.7915,
|
| 1312 |
+
"step": 91500
|
| 1313 |
+
},
|
| 1314 |
+
{
|
| 1315 |
+
"epoch": 2.3694241269187186,
|
| 1316 |
+
"grad_norm": 8.361068725585938,
|
| 1317 |
+
"learning_rate": 6.668432059338622e-06,
|
| 1318 |
+
"loss": 1.774,
|
| 1319 |
+
"step": 92000
|
| 1320 |
+
},
|
| 1321 |
+
{
|
| 1322 |
+
"epoch": 2.38230143195632,
|
| 1323 |
+
"grad_norm": 7.679973602294922,
|
| 1324 |
+
"learning_rate": 6.6666292366333575e-06,
|
| 1325 |
+
"loss": 1.7758,
|
| 1326 |
+
"step": 92500
|
| 1327 |
+
},
|
| 1328 |
+
{
|
| 1329 |
+
"epoch": 2.395178736993922,
|
| 1330 |
+
"grad_norm": 7.885756492614746,
|
| 1331 |
+
"learning_rate": 6.664830019573503e-06,
|
| 1332 |
+
"loss": 1.7808,
|
| 1333 |
+
"step": 93000
|
| 1334 |
+
},
|
| 1335 |
+
{
|
| 1336 |
+
"epoch": 2.408056042031524,
|
| 1337 |
+
"grad_norm": 5.197179794311523,
|
| 1338 |
+
"learning_rate": 6.663027196868239e-06,
|
| 1339 |
+
"loss": 1.7829,
|
| 1340 |
+
"step": 93500
|
| 1341 |
+
},
|
| 1342 |
+
{
|
| 1343 |
+
"epoch": 2.420933347069125,
|
| 1344 |
+
"grad_norm": 5.253746509552002,
|
| 1345 |
+
"learning_rate": 6.661224374162975e-06,
|
| 1346 |
+
"loss": 1.774,
|
| 1347 |
+
"step": 94000
|
| 1348 |
+
},
|
| 1349 |
+
{
|
| 1350 |
+
"epoch": 2.433810652106727,
|
| 1351 |
+
"grad_norm": 6.105677604675293,
|
| 1352 |
+
"learning_rate": 6.659421551457711e-06,
|
| 1353 |
+
"loss": 1.7851,
|
| 1354 |
+
"step": 94500
|
| 1355 |
+
},
|
| 1356 |
+
{
|
| 1357 |
+
"epoch": 2.446687957144329,
|
| 1358 |
+
"grad_norm": 4.1233015060424805,
|
| 1359 |
+
"learning_rate": 6.657618728752447e-06,
|
| 1360 |
+
"loss": 1.77,
|
| 1361 |
+
"step": 95000
|
| 1362 |
+
},
|
| 1363 |
+
{
|
| 1364 |
+
"epoch": 2.4595652621819304,
|
| 1365 |
+
"grad_norm": 5.5101728439331055,
|
| 1366 |
+
"learning_rate": 6.655819511692593e-06,
|
| 1367 |
+
"loss": 1.7782,
|
| 1368 |
+
"step": 95500
|
| 1369 |
+
},
|
| 1370 |
+
{
|
| 1371 |
+
"epoch": 2.4724425672195323,
|
| 1372 |
+
"grad_norm": 6.412257194519043,
|
| 1373 |
+
"learning_rate": 6.654016688987328e-06,
|
| 1374 |
+
"loss": 1.7782,
|
| 1375 |
+
"step": 96000
|
| 1376 |
+
},
|
| 1377 |
+
{
|
| 1378 |
+
"epoch": 2.485319872257134,
|
| 1379 |
+
"grad_norm": 7.24741268157959,
|
| 1380 |
+
"learning_rate": 6.652213866282064e-06,
|
| 1381 |
+
"loss": 1.7836,
|
| 1382 |
+
"step": 96500
|
| 1383 |
+
},
|
| 1384 |
+
{
|
| 1385 |
+
"epoch": 2.4981971772947356,
|
| 1386 |
+
"grad_norm": 4.249999046325684,
|
| 1387 |
+
"learning_rate": 6.650411043576801e-06,
|
| 1388 |
+
"loss": 1.7816,
|
| 1389 |
+
"step": 97000
|
| 1390 |
+
},
|
| 1391 |
+
{
|
| 1392 |
+
"epoch": 2.5110744823323374,
|
| 1393 |
+
"grad_norm": 6.146023750305176,
|
| 1394 |
+
"learning_rate": 6.648608220871536e-06,
|
| 1395 |
+
"loss": 1.7716,
|
| 1396 |
+
"step": 97500
|
| 1397 |
+
},
|
| 1398 |
+
{
|
| 1399 |
+
"epoch": 2.5239517873699393,
|
| 1400 |
+
"grad_norm": 5.631804466247559,
|
| 1401 |
+
"learning_rate": 6.646809003811682e-06,
|
| 1402 |
+
"loss": 1.7912,
|
| 1403 |
+
"step": 98000
|
| 1404 |
+
},
|
| 1405 |
+
{
|
| 1406 |
+
"epoch": 2.536829092407541,
|
| 1407 |
+
"grad_norm": 6.987208843231201,
|
| 1408 |
+
"learning_rate": 6.645006181106418e-06,
|
| 1409 |
+
"loss": 1.7704,
|
| 1410 |
+
"step": 98500
|
| 1411 |
+
},
|
| 1412 |
+
{
|
| 1413 |
+
"epoch": 2.5497063974451426,
|
| 1414 |
+
"grad_norm": 7.857710838317871,
|
| 1415 |
+
"learning_rate": 6.643203358401153e-06,
|
| 1416 |
+
"loss": 1.7756,
|
| 1417 |
+
"step": 99000
|
| 1418 |
+
},
|
| 1419 |
+
{
|
| 1420 |
+
"epoch": 2.5625837024827445,
|
| 1421 |
+
"grad_norm": 6.447986602783203,
|
| 1422 |
+
"learning_rate": 6.64140053569589e-06,
|
| 1423 |
+
"loss": 1.7808,
|
| 1424 |
+
"step": 99500
|
| 1425 |
+
},
|
| 1426 |
+
{
|
| 1427 |
+
"epoch": 2.575461007520346,
|
| 1428 |
+
"grad_norm": 7.811335563659668,
|
| 1429 |
+
"learning_rate": 6.6395977129906255e-06,
|
| 1430 |
+
"loss": 1.7759,
|
| 1431 |
+
"step": 100000
|
| 1432 |
+
},
|
| 1433 |
+
{
|
| 1434 |
+
"epoch": 2.5883383125579478,
|
| 1435 |
+
"grad_norm": 5.068165302276611,
|
| 1436 |
+
"learning_rate": 6.637794890285361e-06,
|
| 1437 |
+
"loss": 1.7931,
|
| 1438 |
+
"step": 100500
|
| 1439 |
+
},
|
| 1440 |
+
{
|
| 1441 |
+
"epoch": 2.6012156175955496,
|
| 1442 |
+
"grad_norm": 7.218021392822266,
|
| 1443 |
+
"learning_rate": 6.635992067580097e-06,
|
| 1444 |
+
"loss": 1.7843,
|
| 1445 |
+
"step": 101000
|
| 1446 |
+
},
|
| 1447 |
+
{
|
| 1448 |
+
"epoch": 2.6140929226331515,
|
| 1449 |
+
"grad_norm": 6.719173908233643,
|
| 1450 |
+
"learning_rate": 6.634192850520243e-06,
|
| 1451 |
+
"loss": 1.7702,
|
| 1452 |
+
"step": 101500
|
| 1453 |
+
},
|
| 1454 |
+
{
|
| 1455 |
+
"epoch": 2.626970227670753,
|
| 1456 |
+
"grad_norm": 5.827301025390625,
|
| 1457 |
+
"learning_rate": 6.632390027814979e-06,
|
| 1458 |
+
"loss": 1.7702,
|
| 1459 |
+
"step": 102000
|
| 1460 |
+
},
|
| 1461 |
+
{
|
| 1462 |
+
"epoch": 2.639847532708355,
|
| 1463 |
+
"grad_norm": 3.2344276905059814,
|
| 1464 |
+
"learning_rate": 6.630587205109715e-06,
|
| 1465 |
+
"loss": 1.7823,
|
| 1466 |
+
"step": 102500
|
| 1467 |
+
},
|
| 1468 |
+
{
|
| 1469 |
+
"epoch": 2.6527248377459567,
|
| 1470 |
+
"grad_norm": 3.8477556705474854,
|
| 1471 |
+
"learning_rate": 6.628784382404451e-06,
|
| 1472 |
+
"loss": 1.7838,
|
| 1473 |
+
"step": 103000
|
| 1474 |
+
},
|
| 1475 |
+
{
|
| 1476 |
+
"epoch": 2.665602142783558,
|
| 1477 |
+
"grad_norm": 5.570420265197754,
|
| 1478 |
+
"learning_rate": 6.626981559699186e-06,
|
| 1479 |
+
"loss": 1.7745,
|
| 1480 |
+
"step": 103500
|
| 1481 |
+
},
|
| 1482 |
+
{
|
| 1483 |
+
"epoch": 2.67847944782116,
|
| 1484 |
+
"grad_norm": 4.79215669631958,
|
| 1485 |
+
"learning_rate": 6.625182342639332e-06,
|
| 1486 |
+
"loss": 1.7727,
|
| 1487 |
+
"step": 104000
|
| 1488 |
+
},
|
| 1489 |
+
{
|
| 1490 |
+
"epoch": 2.691356752858762,
|
| 1491 |
+
"grad_norm": 5.203454494476318,
|
| 1492 |
+
"learning_rate": 6.623379519934068e-06,
|
| 1493 |
+
"loss": 1.7863,
|
| 1494 |
+
"step": 104500
|
| 1495 |
+
},
|
| 1496 |
+
{
|
| 1497 |
+
"epoch": 2.7042340578963633,
|
| 1498 |
+
"grad_norm": 6.587026119232178,
|
| 1499 |
+
"learning_rate": 6.621576697228804e-06,
|
| 1500 |
+
"loss": 1.7687,
|
| 1501 |
+
"step": 105000
|
| 1502 |
+
},
|
| 1503 |
+
{
|
| 1504 |
+
"epoch": 2.717111362933965,
|
| 1505 |
+
"grad_norm": 5.685227394104004,
|
| 1506 |
+
"learning_rate": 6.61977387452354e-06,
|
| 1507 |
+
"loss": 1.7764,
|
| 1508 |
+
"step": 105500
|
| 1509 |
+
},
|
| 1510 |
+
{
|
| 1511 |
+
"epoch": 2.729988667971567,
|
| 1512 |
+
"grad_norm": 5.212521076202393,
|
| 1513 |
+
"learning_rate": 6.6179710518182755e-06,
|
| 1514 |
+
"loss": 1.782,
|
| 1515 |
+
"step": 106000
|
| 1516 |
+
},
|
| 1517 |
+
{
|
| 1518 |
+
"epoch": 2.7428659730091685,
|
| 1519 |
+
"grad_norm": 6.606338024139404,
|
| 1520 |
+
"learning_rate": 6.616171834758422e-06,
|
| 1521 |
+
"loss": 1.7778,
|
| 1522 |
+
"step": 106500
|
| 1523 |
+
},
|
| 1524 |
+
{
|
| 1525 |
+
"epoch": 2.7557432780467703,
|
| 1526 |
+
"grad_norm": 5.185567378997803,
|
| 1527 |
+
"learning_rate": 6.614369012053158e-06,
|
| 1528 |
+
"loss": 1.7819,
|
| 1529 |
+
"step": 107000
|
| 1530 |
+
},
|
| 1531 |
+
{
|
| 1532 |
+
"epoch": 2.768620583084372,
|
| 1533 |
+
"grad_norm": 4.298264503479004,
|
| 1534 |
+
"learning_rate": 6.6125661893478934e-06,
|
| 1535 |
+
"loss": 1.7857,
|
| 1536 |
+
"step": 107500
|
| 1537 |
+
},
|
| 1538 |
+
{
|
| 1539 |
+
"epoch": 2.781497888121974,
|
| 1540 |
+
"grad_norm": 4.79362154006958,
|
| 1541 |
+
"learning_rate": 6.610763366642629e-06,
|
| 1542 |
+
"loss": 1.7912,
|
| 1543 |
+
"step": 108000
|
| 1544 |
+
},
|
| 1545 |
+
{
|
| 1546 |
+
"epoch": 2.7943751931595755,
|
| 1547 |
+
"grad_norm": 4.650264739990234,
|
| 1548 |
+
"learning_rate": 6.608960543937365e-06,
|
| 1549 |
+
"loss": 1.7815,
|
| 1550 |
+
"step": 108500
|
| 1551 |
+
},
|
| 1552 |
+
{
|
| 1553 |
+
"epoch": 2.8072524981971774,
|
| 1554 |
+
"grad_norm": 5.515447616577148,
|
| 1555 |
+
"learning_rate": 6.607161326877511e-06,
|
| 1556 |
+
"loss": 1.7736,
|
| 1557 |
+
"step": 109000
|
| 1558 |
+
},
|
| 1559 |
+
{
|
| 1560 |
+
"epoch": 2.820129803234779,
|
| 1561 |
+
"grad_norm": 4.840721130371094,
|
| 1562 |
+
"learning_rate": 6.605358504172247e-06,
|
| 1563 |
+
"loss": 1.7843,
|
| 1564 |
+
"step": 109500
|
| 1565 |
+
},
|
| 1566 |
+
{
|
| 1567 |
+
"epoch": 2.8330071082723807,
|
| 1568 |
+
"grad_norm": 5.676022052764893,
|
| 1569 |
+
"learning_rate": 6.603555681466983e-06,
|
| 1570 |
+
"loss": 1.7732,
|
| 1571 |
+
"step": 110000
|
| 1572 |
+
},
|
| 1573 |
+
{
|
| 1574 |
+
"epoch": 2.8458844133099825,
|
| 1575 |
+
"grad_norm": 6.819264888763428,
|
| 1576 |
+
"learning_rate": 6.601752858761718e-06,
|
| 1577 |
+
"loss": 1.7786,
|
| 1578 |
+
"step": 110500
|
| 1579 |
+
},
|
| 1580 |
+
{
|
| 1581 |
+
"epoch": 2.8587617183475844,
|
| 1582 |
+
"grad_norm": 5.747689723968506,
|
| 1583 |
+
"learning_rate": 6.599953641701864e-06,
|
| 1584 |
+
"loss": 1.7845,
|
| 1585 |
+
"step": 111000
|
| 1586 |
+
},
|
| 1587 |
+
{
|
| 1588 |
+
"epoch": 2.871639023385186,
|
| 1589 |
+
"grad_norm": 6.419643402099609,
|
| 1590 |
+
"learning_rate": 6.5981508189966e-06,
|
| 1591 |
+
"loss": 1.7745,
|
| 1592 |
+
"step": 111500
|
| 1593 |
+
},
|
| 1594 |
+
{
|
| 1595 |
+
"epoch": 2.8845163284227877,
|
| 1596 |
+
"grad_norm": 5.376995086669922,
|
| 1597 |
+
"learning_rate": 6.596347996291336e-06,
|
| 1598 |
+
"loss": 1.7765,
|
| 1599 |
+
"step": 112000
|
| 1600 |
+
},
|
| 1601 |
+
{
|
| 1602 |
+
"epoch": 2.897393633460389,
|
| 1603 |
+
"grad_norm": 6.216523170471191,
|
| 1604 |
+
"learning_rate": 6.594545173586072e-06,
|
| 1605 |
+
"loss": 1.7718,
|
| 1606 |
+
"step": 112500
|
| 1607 |
+
},
|
| 1608 |
+
{
|
| 1609 |
+
"epoch": 2.910270938497991,
|
| 1610 |
+
"grad_norm": 4.209794998168945,
|
| 1611 |
+
"learning_rate": 6.592742350880808e-06,
|
| 1612 |
+
"loss": 1.7656,
|
| 1613 |
+
"step": 113000
|
| 1614 |
+
},
|
| 1615 |
+
{
|
| 1616 |
+
"epoch": 2.923148243535593,
|
| 1617 |
+
"grad_norm": 4.328191757202148,
|
| 1618 |
+
"learning_rate": 6.590943133820953e-06,
|
| 1619 |
+
"loss": 1.7868,
|
| 1620 |
+
"step": 113500
|
| 1621 |
+
},
|
| 1622 |
+
{
|
| 1623 |
+
"epoch": 2.9360255485731948,
|
| 1624 |
+
"grad_norm": 2.272550344467163,
|
| 1625 |
+
"learning_rate": 6.58914031111569e-06,
|
| 1626 |
+
"loss": 1.7769,
|
| 1627 |
+
"step": 114000
|
| 1628 |
+
},
|
| 1629 |
+
{
|
| 1630 |
+
"epoch": 2.948902853610796,
|
| 1631 |
+
"grad_norm": 4.638970375061035,
|
| 1632 |
+
"learning_rate": 6.587337488410426e-06,
|
| 1633 |
+
"loss": 1.7656,
|
| 1634 |
+
"step": 114500
|
| 1635 |
+
},
|
| 1636 |
+
{
|
| 1637 |
+
"epoch": 2.961780158648398,
|
| 1638 |
+
"grad_norm": 5.238265037536621,
|
| 1639 |
+
"learning_rate": 6.585538271350571e-06,
|
| 1640 |
+
"loss": 1.7628,
|
| 1641 |
+
"step": 115000
|
| 1642 |
+
},
|
| 1643 |
+
{
|
| 1644 |
+
"epoch": 2.974657463686,
|
| 1645 |
+
"grad_norm": 5.025523662567139,
|
| 1646 |
+
"learning_rate": 6.583735448645307e-06,
|
| 1647 |
+
"loss": 1.7875,
|
| 1648 |
+
"step": 115500
|
| 1649 |
+
},
|
| 1650 |
+
{
|
| 1651 |
+
"epoch": 2.9875347687236014,
|
| 1652 |
+
"grad_norm": 5.705627918243408,
|
| 1653 |
+
"learning_rate": 6.581932625940043e-06,
|
| 1654 |
+
"loss": 1.7735,
|
| 1655 |
+
"step": 116000
|
| 1656 |
+
},
|
| 1657 |
+
{
|
| 1658 |
+
"epoch": 3.0,
|
| 1659 |
+
"eval_accuracy": 0.28071705848998996,
|
| 1660 |
+
"eval_f1_macro": 0.15207874587429973,
|
| 1661 |
+
"eval_f1_micro": 0.28071705848998996,
|
| 1662 |
+
"eval_loss": 1.7819397449493408,
|
| 1663 |
+
"eval_runtime": 223.6939,
|
| 1664 |
+
"eval_samples_per_second": 102.493,
|
| 1665 |
+
"eval_steps_per_second": 12.812,
|
| 1666 |
+
"step": 116484
|
| 1667 |
+
}
|
| 1668 |
+
],
|
| 1669 |
+
"logging_steps": 500,
|
| 1670 |
+
"max_steps": 1941400,
|
| 1671 |
+
"num_input_tokens_seen": 0,
|
| 1672 |
+
"num_train_epochs": 50,
|
| 1673 |
+
"save_steps": 500,
|
| 1674 |
+
"stateful_callbacks": {
|
| 1675 |
+
"EarlyStoppingCallback": {
|
| 1676 |
+
"args": {
|
| 1677 |
+
"early_stopping_patience": 3,
|
| 1678 |
+
"early_stopping_threshold": 0.0
|
| 1679 |
+
},
|
| 1680 |
+
"attributes": {
|
| 1681 |
+
"early_stopping_patience_counter": 0
|
| 1682 |
+
}
|
| 1683 |
+
},
|
| 1684 |
+
"TrainerControl": {
|
| 1685 |
+
"args": {
|
| 1686 |
+
"should_epoch_stop": false,
|
| 1687 |
+
"should_evaluate": false,
|
| 1688 |
+
"should_log": false,
|
| 1689 |
+
"should_save": true,
|
| 1690 |
+
"should_training_stop": false
|
| 1691 |
+
},
|
| 1692 |
+
"attributes": {}
|
| 1693 |
+
}
|
| 1694 |
+
},
|
| 1695 |
+
"total_flos": 4.870091010515927e+17,
|
| 1696 |
+
"train_batch_size": 8,
|
| 1697 |
+
"trial_name": null,
|
| 1698 |
+
"trial_params": null
|
| 1699 |
+
}
|
training_args.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:b39f6e65c04bdafa26d7201f33e0101bcdc3fff3e3674275315697d828819750
|
| 3 |
+
size 5368
|