Upload training checkpoints and best model
Browse files- checkpoint-2100/config.json +29 -0
- checkpoint-2100/merges.txt +0 -0
- checkpoint-2100/model.safetensors +3 -0
- checkpoint-2100/optimizer.pt +3 -0
- checkpoint-2100/rng_state.pth +3 -0
- checkpoint-2100/scaler.pt +3 -0
- checkpoint-2100/scheduler.pt +3 -0
- checkpoint-2100/special_tokens_map.json +51 -0
- checkpoint-2100/tokenizer.json +0 -0
- checkpoint-2100/tokenizer_config.json +58 -0
- checkpoint-2100/trainer_state.json +1786 -0
- checkpoint-2100/training_args.bin +3 -0
- checkpoint-2100/vocab.json +0 -0
- checkpoint-2200/config.json +29 -0
- checkpoint-2200/merges.txt +0 -0
- checkpoint-2200/model.safetensors +3 -0
- checkpoint-2200/optimizer.pt +3 -0
- checkpoint-2200/rng_state.pth +3 -0
- checkpoint-2200/scaler.pt +3 -0
- checkpoint-2200/scheduler.pt +3 -0
- checkpoint-2200/special_tokens_map.json +51 -0
- checkpoint-2200/tokenizer.json +0 -0
- checkpoint-2200/tokenizer_config.json +58 -0
- checkpoint-2200/trainer_state.json +1869 -0
- checkpoint-2200/training_args.bin +3 -0
- checkpoint-2200/vocab.json +0 -0
checkpoint-2100/config.json
ADDED
|
@@ -0,0 +1,29 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"architectures": [
|
| 3 |
+
"RobertaForSequenceClassification"
|
| 4 |
+
],
|
| 5 |
+
"attention_probs_dropout_prob": 0.1,
|
| 6 |
+
"bos_token_id": 0,
|
| 7 |
+
"classifier_dropout": null,
|
| 8 |
+
"dtype": "float32",
|
| 9 |
+
"eos_token_id": 2,
|
| 10 |
+
"gradient_checkpointing": false,
|
| 11 |
+
"hidden_act": "gelu",
|
| 12 |
+
"hidden_dropout_prob": 0.1,
|
| 13 |
+
"hidden_size": 768,
|
| 14 |
+
"initializer_range": 0.02,
|
| 15 |
+
"intermediate_size": 3072,
|
| 16 |
+
"layer_norm_eps": 1e-05,
|
| 17 |
+
"max_position_embeddings": 1026,
|
| 18 |
+
"model_type": "roberta",
|
| 19 |
+
"num_attention_heads": 12,
|
| 20 |
+
"num_hidden_layers": 12,
|
| 21 |
+
"output_past": true,
|
| 22 |
+
"pad_token_id": 1,
|
| 23 |
+
"position_embedding_type": "absolute",
|
| 24 |
+
"problem_type": "single_label_classification",
|
| 25 |
+
"transformers_version": "4.56.0",
|
| 26 |
+
"type_vocab_size": 10,
|
| 27 |
+
"use_cache": true,
|
| 28 |
+
"vocab_size": 51416
|
| 29 |
+
}
|
checkpoint-2100/merges.txt
ADDED
|
The diff for this file is too large to render.
See raw diff
|
|
|
checkpoint-2100/model.safetensors
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:3a05ca16ecd06f98a9ceefc2faba95232c23fc27a7b8b2567b8136df172327f6
|
| 3 |
+
size 506505672
|
checkpoint-2100/optimizer.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:18bda20389f78ad6bfa6581ad9adca9bc87acbf2584831e976eefb0e95f7bcd3
|
| 3 |
+
size 1007622411
|
checkpoint-2100/rng_state.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:4c94a304dfdd251f582c2bb1d288755ddac5ef3f7e8f7a1021043c0608bab401
|
| 3 |
+
size 14645
|
checkpoint-2100/scaler.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:12881c5bb3f052d93674ce05408109a22ca56e854712c1e22454903c3cdb1e75
|
| 3 |
+
size 1383
|
checkpoint-2100/scheduler.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:18124ad748bfc730097d810fb9091df86fb381b7b11fa7d08b89a85cb79eb2be
|
| 3 |
+
size 1977
|
checkpoint-2100/special_tokens_map.json
ADDED
|
@@ -0,0 +1,51 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"bos_token": {
|
| 3 |
+
"content": "<s>",
|
| 4 |
+
"lstrip": false,
|
| 5 |
+
"normalized": true,
|
| 6 |
+
"rstrip": false,
|
| 7 |
+
"single_word": false
|
| 8 |
+
},
|
| 9 |
+
"cls_token": {
|
| 10 |
+
"content": "<s>",
|
| 11 |
+
"lstrip": false,
|
| 12 |
+
"normalized": true,
|
| 13 |
+
"rstrip": false,
|
| 14 |
+
"single_word": false
|
| 15 |
+
},
|
| 16 |
+
"eos_token": {
|
| 17 |
+
"content": "</s>",
|
| 18 |
+
"lstrip": false,
|
| 19 |
+
"normalized": true,
|
| 20 |
+
"rstrip": false,
|
| 21 |
+
"single_word": false
|
| 22 |
+
},
|
| 23 |
+
"mask_token": {
|
| 24 |
+
"content": "<mask>",
|
| 25 |
+
"lstrip": true,
|
| 26 |
+
"normalized": true,
|
| 27 |
+
"rstrip": false,
|
| 28 |
+
"single_word": false
|
| 29 |
+
},
|
| 30 |
+
"pad_token": {
|
| 31 |
+
"content": "<pad>",
|
| 32 |
+
"lstrip": false,
|
| 33 |
+
"normalized": true,
|
| 34 |
+
"rstrip": false,
|
| 35 |
+
"single_word": false
|
| 36 |
+
},
|
| 37 |
+
"sep_token": {
|
| 38 |
+
"content": "</s>",
|
| 39 |
+
"lstrip": false,
|
| 40 |
+
"normalized": true,
|
| 41 |
+
"rstrip": false,
|
| 42 |
+
"single_word": false
|
| 43 |
+
},
|
| 44 |
+
"unk_token": {
|
| 45 |
+
"content": "<unk>",
|
| 46 |
+
"lstrip": false,
|
| 47 |
+
"normalized": true,
|
| 48 |
+
"rstrip": false,
|
| 49 |
+
"single_word": false
|
| 50 |
+
}
|
| 51 |
+
}
|
checkpoint-2100/tokenizer.json
ADDED
|
The diff for this file is too large to render.
See raw diff
|
|
|
checkpoint-2100/tokenizer_config.json
ADDED
|
@@ -0,0 +1,58 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"add_prefix_space": false,
|
| 3 |
+
"added_tokens_decoder": {
|
| 4 |
+
"0": {
|
| 5 |
+
"content": "<s>",
|
| 6 |
+
"lstrip": false,
|
| 7 |
+
"normalized": true,
|
| 8 |
+
"rstrip": false,
|
| 9 |
+
"single_word": false,
|
| 10 |
+
"special": true
|
| 11 |
+
},
|
| 12 |
+
"1": {
|
| 13 |
+
"content": "<pad>",
|
| 14 |
+
"lstrip": false,
|
| 15 |
+
"normalized": true,
|
| 16 |
+
"rstrip": false,
|
| 17 |
+
"single_word": false,
|
| 18 |
+
"special": true
|
| 19 |
+
},
|
| 20 |
+
"2": {
|
| 21 |
+
"content": "</s>",
|
| 22 |
+
"lstrip": false,
|
| 23 |
+
"normalized": true,
|
| 24 |
+
"rstrip": false,
|
| 25 |
+
"single_word": false,
|
| 26 |
+
"special": true
|
| 27 |
+
},
|
| 28 |
+
"3": {
|
| 29 |
+
"content": "<unk>",
|
| 30 |
+
"lstrip": false,
|
| 31 |
+
"normalized": true,
|
| 32 |
+
"rstrip": false,
|
| 33 |
+
"single_word": false,
|
| 34 |
+
"special": true
|
| 35 |
+
},
|
| 36 |
+
"4": {
|
| 37 |
+
"content": "<mask>",
|
| 38 |
+
"lstrip": true,
|
| 39 |
+
"normalized": true,
|
| 40 |
+
"rstrip": false,
|
| 41 |
+
"single_word": false,
|
| 42 |
+
"special": true
|
| 43 |
+
}
|
| 44 |
+
},
|
| 45 |
+
"bos_token": "<s>",
|
| 46 |
+
"clean_up_tokenization_spaces": false,
|
| 47 |
+
"cls_token": "<s>",
|
| 48 |
+
"eos_token": "</s>",
|
| 49 |
+
"errors": "replace",
|
| 50 |
+
"extra_special_tokens": {},
|
| 51 |
+
"mask_token": "<mask>",
|
| 52 |
+
"model_max_length": 1000000000000000019884624838656,
|
| 53 |
+
"pad_token": "<pad>",
|
| 54 |
+
"sep_token": "</s>",
|
| 55 |
+
"tokenizer_class": "RobertaTokenizer",
|
| 56 |
+
"trim_offsets": true,
|
| 57 |
+
"unk_token": "<unk>"
|
| 58 |
+
}
|
checkpoint-2100/trainer_state.json
ADDED
|
@@ -0,0 +1,1786 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_global_step": 2100,
|
| 3 |
+
"best_metric": 0.9776543362648145,
|
| 4 |
+
"best_model_checkpoint": "taskA-unixcoder-focal/checkpoint-2100",
|
| 5 |
+
"epoch": 0.25694359476324485,
|
| 6 |
+
"eval_steps": 100,
|
| 7 |
+
"global_step": 2100,
|
| 8 |
+
"is_hyper_param_search": false,
|
| 9 |
+
"is_local_process_zero": true,
|
| 10 |
+
"is_world_process_zero": true,
|
| 11 |
+
"log_history": [
|
| 12 |
+
{
|
| 13 |
+
"epoch": 0.001223540927444023,
|
| 14 |
+
"grad_norm": 0.1898110806941986,
|
| 15 |
+
"learning_rate": 3.6734693877551025e-07,
|
| 16 |
+
"loss": 0.4422,
|
| 17 |
+
"step": 10
|
| 18 |
+
},
|
| 19 |
+
{
|
| 20 |
+
"epoch": 0.002447081854888046,
|
| 21 |
+
"grad_norm": 0.3350900411605835,
|
| 22 |
+
"learning_rate": 7.346938775510205e-07,
|
| 23 |
+
"loss": 0.4323,
|
| 24 |
+
"step": 20
|
| 25 |
+
},
|
| 26 |
+
{
|
| 27 |
+
"epoch": 0.003670622782332069,
|
| 28 |
+
"grad_norm": 0.3723449110984802,
|
| 29 |
+
"learning_rate": 1.142857142857143e-06,
|
| 30 |
+
"loss": 0.4362,
|
| 31 |
+
"step": 30
|
| 32 |
+
},
|
| 33 |
+
{
|
| 34 |
+
"epoch": 0.004894163709776092,
|
| 35 |
+
"grad_norm": 0.3299769461154938,
|
| 36 |
+
"learning_rate": 1.5510204081632654e-06,
|
| 37 |
+
"loss": 0.4191,
|
| 38 |
+
"step": 40
|
| 39 |
+
},
|
| 40 |
+
{
|
| 41 |
+
"epoch": 0.006117704637220115,
|
| 42 |
+
"grad_norm": 0.26681044697761536,
|
| 43 |
+
"learning_rate": 1.9591836734693877e-06,
|
| 44 |
+
"loss": 0.4289,
|
| 45 |
+
"step": 50
|
| 46 |
+
},
|
| 47 |
+
{
|
| 48 |
+
"epoch": 0.007341245564664138,
|
| 49 |
+
"grad_norm": 0.30233338475227356,
|
| 50 |
+
"learning_rate": 2.3673469387755103e-06,
|
| 51 |
+
"loss": 0.4083,
|
| 52 |
+
"step": 60
|
| 53 |
+
},
|
| 54 |
+
{
|
| 55 |
+
"epoch": 0.008564786492108161,
|
| 56 |
+
"grad_norm": 0.31031334400177,
|
| 57 |
+
"learning_rate": 2.775510204081633e-06,
|
| 58 |
+
"loss": 0.404,
|
| 59 |
+
"step": 70
|
| 60 |
+
},
|
| 61 |
+
{
|
| 62 |
+
"epoch": 0.009788327419552184,
|
| 63 |
+
"grad_norm": 0.2552475035190582,
|
| 64 |
+
"learning_rate": 3.183673469387755e-06,
|
| 65 |
+
"loss": 0.3889,
|
| 66 |
+
"step": 80
|
| 67 |
+
},
|
| 68 |
+
{
|
| 69 |
+
"epoch": 0.011011868346996207,
|
| 70 |
+
"grad_norm": 0.3131052553653717,
|
| 71 |
+
"learning_rate": 3.5918367346938777e-06,
|
| 72 |
+
"loss": 0.3813,
|
| 73 |
+
"step": 90
|
| 74 |
+
},
|
| 75 |
+
{
|
| 76 |
+
"epoch": 0.01223540927444023,
|
| 77 |
+
"grad_norm": 0.3377229571342468,
|
| 78 |
+
"learning_rate": 4.000000000000001e-06,
|
| 79 |
+
"loss": 0.3727,
|
| 80 |
+
"step": 100
|
| 81 |
+
},
|
| 82 |
+
{
|
| 83 |
+
"epoch": 0.01223540927444023,
|
| 84 |
+
"eval_accuracy": 0.6961,
|
| 85 |
+
"eval_f1_weighted": 0.6955947095331424,
|
| 86 |
+
"eval_loss": 0.15026023983955383,
|
| 87 |
+
"eval_macro_f1": 0.6959280508399847,
|
| 88 |
+
"eval_precision": 0.7019396279846728,
|
| 89 |
+
"eval_recall": 0.6961,
|
| 90 |
+
"eval_runtime": 789.7648,
|
| 91 |
+
"eval_samples_per_second": 126.62,
|
| 92 |
+
"eval_steps_per_second": 0.99,
|
| 93 |
+
"step": 100
|
| 94 |
+
},
|
| 95 |
+
{
|
| 96 |
+
"epoch": 0.013458950201884253,
|
| 97 |
+
"grad_norm": 0.3366774618625641,
|
| 98 |
+
"learning_rate": 4.408163265306123e-06,
|
| 99 |
+
"loss": 0.3671,
|
| 100 |
+
"step": 110
|
| 101 |
+
},
|
| 102 |
+
{
|
| 103 |
+
"epoch": 0.014682491129328276,
|
| 104 |
+
"grad_norm": 0.33491575717926025,
|
| 105 |
+
"learning_rate": 4.816326530612245e-06,
|
| 106 |
+
"loss": 0.3498,
|
| 107 |
+
"step": 120
|
| 108 |
+
},
|
| 109 |
+
{
|
| 110 |
+
"epoch": 0.0159060320567723,
|
| 111 |
+
"grad_norm": 0.4119006097316742,
|
| 112 |
+
"learning_rate": 5.2244897959183684e-06,
|
| 113 |
+
"loss": 0.3409,
|
| 114 |
+
"step": 130
|
| 115 |
+
},
|
| 116 |
+
{
|
| 117 |
+
"epoch": 0.017129572984216322,
|
| 118 |
+
"grad_norm": 0.2964725196361542,
|
| 119 |
+
"learning_rate": 5.63265306122449e-06,
|
| 120 |
+
"loss": 0.3433,
|
| 121 |
+
"step": 140
|
| 122 |
+
},
|
| 123 |
+
{
|
| 124 |
+
"epoch": 0.018353113911660345,
|
| 125 |
+
"grad_norm": 0.3075334429740906,
|
| 126 |
+
"learning_rate": 6.040816326530613e-06,
|
| 127 |
+
"loss": 0.3298,
|
| 128 |
+
"step": 150
|
| 129 |
+
},
|
| 130 |
+
{
|
| 131 |
+
"epoch": 0.019576654839104368,
|
| 132 |
+
"grad_norm": 0.33718687295913696,
|
| 133 |
+
"learning_rate": 6.4489795918367345e-06,
|
| 134 |
+
"loss": 0.3297,
|
| 135 |
+
"step": 160
|
| 136 |
+
},
|
| 137 |
+
{
|
| 138 |
+
"epoch": 0.02080019576654839,
|
| 139 |
+
"grad_norm": 0.2809618413448334,
|
| 140 |
+
"learning_rate": 6.857142857142858e-06,
|
| 141 |
+
"loss": 0.3275,
|
| 142 |
+
"step": 170
|
| 143 |
+
},
|
| 144 |
+
{
|
| 145 |
+
"epoch": 0.022023736693992414,
|
| 146 |
+
"grad_norm": 0.37805014848709106,
|
| 147 |
+
"learning_rate": 7.2653061224489805e-06,
|
| 148 |
+
"loss": 0.3201,
|
| 149 |
+
"step": 180
|
| 150 |
+
},
|
| 151 |
+
{
|
| 152 |
+
"epoch": 0.023247277621436437,
|
| 153 |
+
"grad_norm": 0.2940429151058197,
|
| 154 |
+
"learning_rate": 7.673469387755102e-06,
|
| 155 |
+
"loss": 0.3042,
|
| 156 |
+
"step": 190
|
| 157 |
+
},
|
| 158 |
+
{
|
| 159 |
+
"epoch": 0.02447081854888046,
|
| 160 |
+
"grad_norm": 0.3286060690879822,
|
| 161 |
+
"learning_rate": 8.081632653061225e-06,
|
| 162 |
+
"loss": 0.3071,
|
| 163 |
+
"step": 200
|
| 164 |
+
},
|
| 165 |
+
{
|
| 166 |
+
"epoch": 0.02447081854888046,
|
| 167 |
+
"eval_accuracy": 0.77807,
|
| 168 |
+
"eval_f1_weighted": 0.7781129395311969,
|
| 169 |
+
"eval_loss": 0.12153773754835129,
|
| 170 |
+
"eval_macro_f1": 0.7780651598452646,
|
| 171 |
+
"eval_precision": 0.7808306361972198,
|
| 172 |
+
"eval_recall": 0.77807,
|
| 173 |
+
"eval_runtime": 792.7157,
|
| 174 |
+
"eval_samples_per_second": 126.149,
|
| 175 |
+
"eval_steps_per_second": 0.986,
|
| 176 |
+
"step": 200
|
| 177 |
+
},
|
| 178 |
+
{
|
| 179 |
+
"epoch": 0.025694359476324483,
|
| 180 |
+
"grad_norm": 0.33857154846191406,
|
| 181 |
+
"learning_rate": 8.489795918367347e-06,
|
| 182 |
+
"loss": 0.2964,
|
| 183 |
+
"step": 210
|
| 184 |
+
},
|
| 185 |
+
{
|
| 186 |
+
"epoch": 0.026917900403768506,
|
| 187 |
+
"grad_norm": 0.2752092182636261,
|
| 188 |
+
"learning_rate": 8.89795918367347e-06,
|
| 189 |
+
"loss": 0.3,
|
| 190 |
+
"step": 220
|
| 191 |
+
},
|
| 192 |
+
{
|
| 193 |
+
"epoch": 0.02814144133121253,
|
| 194 |
+
"grad_norm": 0.4952544569969177,
|
| 195 |
+
"learning_rate": 9.306122448979593e-06,
|
| 196 |
+
"loss": 0.294,
|
| 197 |
+
"step": 230
|
| 198 |
+
},
|
| 199 |
+
{
|
| 200 |
+
"epoch": 0.029364982258656552,
|
| 201 |
+
"grad_norm": 0.37926042079925537,
|
| 202 |
+
"learning_rate": 9.714285714285715e-06,
|
| 203 |
+
"loss": 0.2964,
|
| 204 |
+
"step": 240
|
| 205 |
+
},
|
| 206 |
+
{
|
| 207 |
+
"epoch": 0.030588523186100575,
|
| 208 |
+
"grad_norm": 0.3222455680370331,
|
| 209 |
+
"learning_rate": 1.0122448979591836e-05,
|
| 210 |
+
"loss": 0.283,
|
| 211 |
+
"step": 250
|
| 212 |
+
},
|
| 213 |
+
{
|
| 214 |
+
"epoch": 0.0318120641135446,
|
| 215 |
+
"grad_norm": 0.2930658459663391,
|
| 216 |
+
"learning_rate": 1.053061224489796e-05,
|
| 217 |
+
"loss": 0.2793,
|
| 218 |
+
"step": 260
|
| 219 |
+
},
|
| 220 |
+
{
|
| 221 |
+
"epoch": 0.03303560504098862,
|
| 222 |
+
"grad_norm": 0.4313414692878723,
|
| 223 |
+
"learning_rate": 1.0938775510204083e-05,
|
| 224 |
+
"loss": 0.2728,
|
| 225 |
+
"step": 270
|
| 226 |
+
},
|
| 227 |
+
{
|
| 228 |
+
"epoch": 0.034259145968432644,
|
| 229 |
+
"grad_norm": 0.37351569533348083,
|
| 230 |
+
"learning_rate": 1.1346938775510206e-05,
|
| 231 |
+
"loss": 0.2752,
|
| 232 |
+
"step": 280
|
| 233 |
+
},
|
| 234 |
+
{
|
| 235 |
+
"epoch": 0.035482686895876664,
|
| 236 |
+
"grad_norm": 0.22772285342216492,
|
| 237 |
+
"learning_rate": 1.1755102040816328e-05,
|
| 238 |
+
"loss": 0.274,
|
| 239 |
+
"step": 290
|
| 240 |
+
},
|
| 241 |
+
{
|
| 242 |
+
"epoch": 0.03670622782332069,
|
| 243 |
+
"grad_norm": 0.15888753533363342,
|
| 244 |
+
"learning_rate": 1.2163265306122449e-05,
|
| 245 |
+
"loss": 0.264,
|
| 246 |
+
"step": 300
|
| 247 |
+
},
|
| 248 |
+
{
|
| 249 |
+
"epoch": 0.03670622782332069,
|
| 250 |
+
"eval_accuracy": 0.82326,
|
| 251 |
+
"eval_f1_weighted": 0.8227034206790359,
|
| 252 |
+
"eval_loss": 0.10234300792217255,
|
| 253 |
+
"eval_macro_f1": 0.8230100448030374,
|
| 254 |
+
"eval_precision": 0.8338333810964687,
|
| 255 |
+
"eval_recall": 0.82326,
|
| 256 |
+
"eval_runtime": 789.0168,
|
| 257 |
+
"eval_samples_per_second": 126.74,
|
| 258 |
+
"eval_steps_per_second": 0.991,
|
| 259 |
+
"step": 300
|
| 260 |
+
},
|
| 261 |
+
{
|
| 262 |
+
"epoch": 0.03792976875076471,
|
| 263 |
+
"grad_norm": 0.23560188710689545,
|
| 264 |
+
"learning_rate": 1.2571428571428572e-05,
|
| 265 |
+
"loss": 0.2753,
|
| 266 |
+
"step": 310
|
| 267 |
+
},
|
| 268 |
+
{
|
| 269 |
+
"epoch": 0.039153309678208736,
|
| 270 |
+
"grad_norm": 0.18699614703655243,
|
| 271 |
+
"learning_rate": 1.2979591836734696e-05,
|
| 272 |
+
"loss": 0.2565,
|
| 273 |
+
"step": 320
|
| 274 |
+
},
|
| 275 |
+
{
|
| 276 |
+
"epoch": 0.040376850605652756,
|
| 277 |
+
"grad_norm": 0.24035991728305817,
|
| 278 |
+
"learning_rate": 1.3387755102040818e-05,
|
| 279 |
+
"loss": 0.2716,
|
| 280 |
+
"step": 330
|
| 281 |
+
},
|
| 282 |
+
{
|
| 283 |
+
"epoch": 0.04160039153309678,
|
| 284 |
+
"grad_norm": 0.15309660136699677,
|
| 285 |
+
"learning_rate": 1.3795918367346941e-05,
|
| 286 |
+
"loss": 0.2515,
|
| 287 |
+
"step": 340
|
| 288 |
+
},
|
| 289 |
+
{
|
| 290 |
+
"epoch": 0.0428239324605408,
|
| 291 |
+
"grad_norm": 0.3382055461406708,
|
| 292 |
+
"learning_rate": 1.4204081632653062e-05,
|
| 293 |
+
"loss": 0.2597,
|
| 294 |
+
"step": 350
|
| 295 |
+
},
|
| 296 |
+
{
|
| 297 |
+
"epoch": 0.04404747338798483,
|
| 298 |
+
"grad_norm": 0.24439455568790436,
|
| 299 |
+
"learning_rate": 1.4612244897959185e-05,
|
| 300 |
+
"loss": 0.247,
|
| 301 |
+
"step": 360
|
| 302 |
+
},
|
| 303 |
+
{
|
| 304 |
+
"epoch": 0.04527101431542885,
|
| 305 |
+
"grad_norm": 0.26233720779418945,
|
| 306 |
+
"learning_rate": 1.5020408163265307e-05,
|
| 307 |
+
"loss": 0.2431,
|
| 308 |
+
"step": 370
|
| 309 |
+
},
|
| 310 |
+
{
|
| 311 |
+
"epoch": 0.046494555242872875,
|
| 312 |
+
"grad_norm": 0.20477451384067535,
|
| 313 |
+
"learning_rate": 1.5387755102040818e-05,
|
| 314 |
+
"loss": 0.2538,
|
| 315 |
+
"step": 380
|
| 316 |
+
},
|
| 317 |
+
{
|
| 318 |
+
"epoch": 0.047718096170316894,
|
| 319 |
+
"grad_norm": 0.2765439748764038,
|
| 320 |
+
"learning_rate": 1.5795918367346942e-05,
|
| 321 |
+
"loss": 0.2404,
|
| 322 |
+
"step": 390
|
| 323 |
+
},
|
| 324 |
+
{
|
| 325 |
+
"epoch": 0.04894163709776092,
|
| 326 |
+
"grad_norm": 0.13196279108524323,
|
| 327 |
+
"learning_rate": 1.6204081632653063e-05,
|
| 328 |
+
"loss": 0.2393,
|
| 329 |
+
"step": 400
|
| 330 |
+
},
|
| 331 |
+
{
|
| 332 |
+
"epoch": 0.04894163709776092,
|
| 333 |
+
"eval_accuracy": 0.84563,
|
| 334 |
+
"eval_f1_weighted": 0.8448128964509274,
|
| 335 |
+
"eval_loss": 0.09028046578168869,
|
| 336 |
+
"eval_macro_f1": 0.8451923512067675,
|
| 337 |
+
"eval_precision": 0.861024732271907,
|
| 338 |
+
"eval_recall": 0.84563,
|
| 339 |
+
"eval_runtime": 787.6067,
|
| 340 |
+
"eval_samples_per_second": 126.967,
|
| 341 |
+
"eval_steps_per_second": 0.993,
|
| 342 |
+
"step": 400
|
| 343 |
+
},
|
| 344 |
+
{
|
| 345 |
+
"epoch": 0.05016517802520494,
|
| 346 |
+
"grad_norm": 0.0924634337425232,
|
| 347 |
+
"learning_rate": 1.6612244897959184e-05,
|
| 348 |
+
"loss": 0.2331,
|
| 349 |
+
"step": 410
|
| 350 |
+
},
|
| 351 |
+
{
|
| 352 |
+
"epoch": 0.05138871895264897,
|
| 353 |
+
"grad_norm": 0.277784138917923,
|
| 354 |
+
"learning_rate": 1.7020408163265308e-05,
|
| 355 |
+
"loss": 0.2301,
|
| 356 |
+
"step": 420
|
| 357 |
+
},
|
| 358 |
+
{
|
| 359 |
+
"epoch": 0.052612259880092986,
|
| 360 |
+
"grad_norm": 0.32299959659576416,
|
| 361 |
+
"learning_rate": 1.742857142857143e-05,
|
| 362 |
+
"loss": 0.2218,
|
| 363 |
+
"step": 430
|
| 364 |
+
},
|
| 365 |
+
{
|
| 366 |
+
"epoch": 0.05383580080753701,
|
| 367 |
+
"grad_norm": 0.18170586228370667,
|
| 368 |
+
"learning_rate": 1.7836734693877553e-05,
|
| 369 |
+
"loss": 0.2318,
|
| 370 |
+
"step": 440
|
| 371 |
+
},
|
| 372 |
+
{
|
| 373 |
+
"epoch": 0.05505934173498103,
|
| 374 |
+
"grad_norm": 0.1520216166973114,
|
| 375 |
+
"learning_rate": 1.8244897959183674e-05,
|
| 376 |
+
"loss": 0.2136,
|
| 377 |
+
"step": 450
|
| 378 |
+
},
|
| 379 |
+
{
|
| 380 |
+
"epoch": 0.05628288266242506,
|
| 381 |
+
"grad_norm": 0.23452645540237427,
|
| 382 |
+
"learning_rate": 1.8653061224489795e-05,
|
| 383 |
+
"loss": 0.2219,
|
| 384 |
+
"step": 460
|
| 385 |
+
},
|
| 386 |
+
{
|
| 387 |
+
"epoch": 0.05750642358986908,
|
| 388 |
+
"grad_norm": 0.10036301612854004,
|
| 389 |
+
"learning_rate": 1.906122448979592e-05,
|
| 390 |
+
"loss": 0.2213,
|
| 391 |
+
"step": 470
|
| 392 |
+
},
|
| 393 |
+
{
|
| 394 |
+
"epoch": 0.058729964517313105,
|
| 395 |
+
"grad_norm": 0.24100220203399658,
|
| 396 |
+
"learning_rate": 1.9469387755102044e-05,
|
| 397 |
+
"loss": 0.2079,
|
| 398 |
+
"step": 480
|
| 399 |
+
},
|
| 400 |
+
{
|
| 401 |
+
"epoch": 0.059953505444757124,
|
| 402 |
+
"grad_norm": 0.28893032670021057,
|
| 403 |
+
"learning_rate": 1.9877551020408165e-05,
|
| 404 |
+
"loss": 0.1978,
|
| 405 |
+
"step": 490
|
| 406 |
+
},
|
| 407 |
+
{
|
| 408 |
+
"epoch": 0.06117704637220115,
|
| 409 |
+
"grad_norm": 0.22918537259101868,
|
| 410 |
+
"learning_rate": 1.9999836143766126e-05,
|
| 411 |
+
"loss": 0.202,
|
| 412 |
+
"step": 500
|
| 413 |
+
},
|
| 414 |
+
{
|
| 415 |
+
"epoch": 0.06117704637220115,
|
| 416 |
+
"eval_accuracy": 0.91628,
|
| 417 |
+
"eval_f1_weighted": 0.9163231793856368,
|
| 418 |
+
"eval_loss": 0.055908624082803726,
|
| 419 |
+
"eval_macro_f1": 0.9162187423616152,
|
| 420 |
+
"eval_precision": 0.9170320825159874,
|
| 421 |
+
"eval_recall": 0.91628,
|
| 422 |
+
"eval_runtime": 760.1138,
|
| 423 |
+
"eval_samples_per_second": 131.559,
|
| 424 |
+
"eval_steps_per_second": 1.029,
|
| 425 |
+
"step": 500
|
| 426 |
+
},
|
| 427 |
+
{
|
| 428 |
+
"epoch": 0.06240058729964517,
|
| 429 |
+
"grad_norm": 0.2930404841899872,
|
| 430 |
+
"learning_rate": 1.999903359554741e-05,
|
| 431 |
+
"loss": 0.1979,
|
| 432 |
+
"step": 510
|
| 433 |
+
},
|
| 434 |
+
{
|
| 435 |
+
"epoch": 0.0636241282270892,
|
| 436 |
+
"grad_norm": 0.0889664739370346,
|
| 437 |
+
"learning_rate": 1.9997562312908153e-05,
|
| 438 |
+
"loss": 0.1953,
|
| 439 |
+
"step": 520
|
| 440 |
+
},
|
| 441 |
+
{
|
| 442 |
+
"epoch": 0.06484766915453322,
|
| 443 |
+
"grad_norm": 0.1687588095664978,
|
| 444 |
+
"learning_rate": 1.9995422394247598e-05,
|
| 445 |
+
"loss": 0.1935,
|
| 446 |
+
"step": 530
|
| 447 |
+
},
|
| 448 |
+
{
|
| 449 |
+
"epoch": 0.06607121008197724,
|
| 450 |
+
"grad_norm": 0.2399725466966629,
|
| 451 |
+
"learning_rate": 1.9992613982683296e-05,
|
| 452 |
+
"loss": 0.1818,
|
| 453 |
+
"step": 540
|
| 454 |
+
},
|
| 455 |
+
{
|
| 456 |
+
"epoch": 0.06729475100942127,
|
| 457 |
+
"grad_norm": 0.2016785889863968,
|
| 458 |
+
"learning_rate": 1.9989137266041552e-05,
|
| 459 |
+
"loss": 0.1893,
|
| 460 |
+
"step": 550
|
| 461 |
+
},
|
| 462 |
+
{
|
| 463 |
+
"epoch": 0.06851829193686529,
|
| 464 |
+
"grad_norm": 0.17420713603496552,
|
| 465 |
+
"learning_rate": 1.9984992476844845e-05,
|
| 466 |
+
"loss": 0.1895,
|
| 467 |
+
"step": 560
|
| 468 |
+
},
|
| 469 |
+
{
|
| 470 |
+
"epoch": 0.06974183286430931,
|
| 471 |
+
"grad_norm": 0.2180609554052353,
|
| 472 |
+
"learning_rate": 1.998017989229629e-05,
|
| 473 |
+
"loss": 0.1836,
|
| 474 |
+
"step": 570
|
| 475 |
+
},
|
| 476 |
+
{
|
| 477 |
+
"epoch": 0.07096537379175333,
|
| 478 |
+
"grad_norm": 0.15563039481639862,
|
| 479 |
+
"learning_rate": 1.9974699834261074e-05,
|
| 480 |
+
"loss": 0.182,
|
| 481 |
+
"step": 580
|
| 482 |
+
},
|
| 483 |
+
{
|
| 484 |
+
"epoch": 0.07218891471919736,
|
| 485 |
+
"grad_norm": 0.25410088896751404,
|
| 486 |
+
"learning_rate": 1.996855266924497e-05,
|
| 487 |
+
"loss": 0.1862,
|
| 488 |
+
"step": 590
|
| 489 |
+
},
|
| 490 |
+
{
|
| 491 |
+
"epoch": 0.07341245564664138,
|
| 492 |
+
"grad_norm": 0.24206259846687317,
|
| 493 |
+
"learning_rate": 1.996173880836979e-05,
|
| 494 |
+
"loss": 0.1818,
|
| 495 |
+
"step": 600
|
| 496 |
+
},
|
| 497 |
+
{
|
| 498 |
+
"epoch": 0.07341245564664138,
|
| 499 |
+
"eval_accuracy": 0.91633,
|
| 500 |
+
"eval_f1_weighted": 0.9162437523349878,
|
| 501 |
+
"eval_loss": 0.05625538155436516,
|
| 502 |
+
"eval_macro_f1": 0.9163072974421806,
|
| 503 |
+
"eval_precision": 0.9233764457540103,
|
| 504 |
+
"eval_recall": 0.91633,
|
| 505 |
+
"eval_runtime": 759.6815,
|
| 506 |
+
"eval_samples_per_second": 131.634,
|
| 507 |
+
"eval_steps_per_second": 1.029,
|
| 508 |
+
"step": 600
|
| 509 |
+
},
|
| 510 |
+
{
|
| 511 |
+
"epoch": 0.0746359965740854,
|
| 512 |
+
"grad_norm": 0.27251410484313965,
|
| 513 |
+
"learning_rate": 1.9954258707345903e-05,
|
| 514 |
+
"loss": 0.1845,
|
| 515 |
+
"step": 610
|
| 516 |
+
},
|
| 517 |
+
{
|
| 518 |
+
"epoch": 0.07585953750152942,
|
| 519 |
+
"grad_norm": 0.1545839011669159,
|
| 520 |
+
"learning_rate": 1.9946112866441757e-05,
|
| 521 |
+
"loss": 0.1861,
|
| 522 |
+
"step": 620
|
| 523 |
+
},
|
| 524 |
+
{
|
| 525 |
+
"epoch": 0.07708307842897345,
|
| 526 |
+
"grad_norm": 0.2535873055458069,
|
| 527 |
+
"learning_rate": 1.9937301830450423e-05,
|
| 528 |
+
"loss": 0.1772,
|
| 529 |
+
"step": 630
|
| 530 |
+
},
|
| 531 |
+
{
|
| 532 |
+
"epoch": 0.07830661935641747,
|
| 533 |
+
"grad_norm": 0.2617948651313782,
|
| 534 |
+
"learning_rate": 1.992782618865315e-05,
|
| 535 |
+
"loss": 0.1711,
|
| 536 |
+
"step": 640
|
| 537 |
+
},
|
| 538 |
+
{
|
| 539 |
+
"epoch": 0.07953016028386149,
|
| 540 |
+
"grad_norm": 0.38676971197128296,
|
| 541 |
+
"learning_rate": 1.9917686574779972e-05,
|
| 542 |
+
"loss": 0.1748,
|
| 543 |
+
"step": 650
|
| 544 |
+
},
|
| 545 |
+
{
|
| 546 |
+
"epoch": 0.08075370121130551,
|
| 547 |
+
"grad_norm": 0.1511957049369812,
|
| 548 |
+
"learning_rate": 1.9906883666967296e-05,
|
| 549 |
+
"loss": 0.1825,
|
| 550 |
+
"step": 660
|
| 551 |
+
},
|
| 552 |
+
{
|
| 553 |
+
"epoch": 0.08197724213874955,
|
| 554 |
+
"grad_norm": 0.2525976002216339,
|
| 555 |
+
"learning_rate": 1.989541818771258e-05,
|
| 556 |
+
"loss": 0.1651,
|
| 557 |
+
"step": 670
|
| 558 |
+
},
|
| 559 |
+
{
|
| 560 |
+
"epoch": 0.08320078306619356,
|
| 561 |
+
"grad_norm": 0.25827622413635254,
|
| 562 |
+
"learning_rate": 1.9883290903825994e-05,
|
| 563 |
+
"loss": 0.1763,
|
| 564 |
+
"step": 680
|
| 565 |
+
},
|
| 566 |
+
{
|
| 567 |
+
"epoch": 0.08442432399363758,
|
| 568 |
+
"grad_norm": 0.24877820909023285,
|
| 569 |
+
"learning_rate": 1.9870502626379127e-05,
|
| 570 |
+
"loss": 0.1652,
|
| 571 |
+
"step": 690
|
| 572 |
+
},
|
| 573 |
+
{
|
| 574 |
+
"epoch": 0.0856478649210816,
|
| 575 |
+
"grad_norm": 0.278772234916687,
|
| 576 |
+
"learning_rate": 1.9857054210650776e-05,
|
| 577 |
+
"loss": 0.1651,
|
| 578 |
+
"step": 700
|
| 579 |
+
},
|
| 580 |
+
{
|
| 581 |
+
"epoch": 0.0856478649210816,
|
| 582 |
+
"eval_accuracy": 0.94905,
|
| 583 |
+
"eval_f1_weighted": 0.9490731118481598,
|
| 584 |
+
"eval_loss": 0.03884471580386162,
|
| 585 |
+
"eval_macro_f1": 0.9489981325237288,
|
| 586 |
+
"eval_precision": 0.9494795831024372,
|
| 587 |
+
"eval_recall": 0.94905,
|
| 588 |
+
"eval_runtime": 759.9337,
|
| 589 |
+
"eval_samples_per_second": 131.59,
|
| 590 |
+
"eval_steps_per_second": 1.029,
|
| 591 |
+
"step": 700
|
| 592 |
+
},
|
| 593 |
+
{
|
| 594 |
+
"epoch": 0.08687140584852564,
|
| 595 |
+
"grad_norm": 0.1895805150270462,
|
| 596 |
+
"learning_rate": 1.9842946556069705e-05,
|
| 597 |
+
"loss": 0.1628,
|
| 598 |
+
"step": 710
|
| 599 |
+
},
|
| 600 |
+
{
|
| 601 |
+
"epoch": 0.08809494677596966,
|
| 602 |
+
"grad_norm": 0.1171497106552124,
|
| 603 |
+
"learning_rate": 1.982818060615452e-05,
|
| 604 |
+
"loss": 0.1715,
|
| 605 |
+
"step": 720
|
| 606 |
+
},
|
| 607 |
+
{
|
| 608 |
+
"epoch": 0.08931848770341368,
|
| 609 |
+
"grad_norm": 0.22032728791236877,
|
| 610 |
+
"learning_rate": 1.9812757348450547e-05,
|
| 611 |
+
"loss": 0.1712,
|
| 612 |
+
"step": 730
|
| 613 |
+
},
|
| 614 |
+
{
|
| 615 |
+
"epoch": 0.0905420286308577,
|
| 616 |
+
"grad_norm": 0.20700488984584808,
|
| 617 |
+
"learning_rate": 1.9796677814463812e-05,
|
| 618 |
+
"loss": 0.1638,
|
| 619 |
+
"step": 740
|
| 620 |
+
},
|
| 621 |
+
{
|
| 622 |
+
"epoch": 0.09176556955830173,
|
| 623 |
+
"grad_norm": 0.24688613414764404,
|
| 624 |
+
"learning_rate": 1.977994307959202e-05,
|
| 625 |
+
"loss": 0.1647,
|
| 626 |
+
"step": 750
|
| 627 |
+
},
|
| 628 |
+
{
|
| 629 |
+
"epoch": 0.09298911048574575,
|
| 630 |
+
"grad_norm": 0.1886432021856308,
|
| 631 |
+
"learning_rate": 1.9762554263052655e-05,
|
| 632 |
+
"loss": 0.1734,
|
| 633 |
+
"step": 760
|
| 634 |
+
},
|
| 635 |
+
{
|
| 636 |
+
"epoch": 0.09421265141318977,
|
| 637 |
+
"grad_norm": 0.08407370001077652,
|
| 638 |
+
"learning_rate": 1.9744512527808127e-05,
|
| 639 |
+
"loss": 0.1755,
|
| 640 |
+
"step": 770
|
| 641 |
+
},
|
| 642 |
+
{
|
| 643 |
+
"epoch": 0.09543619234063379,
|
| 644 |
+
"grad_norm": 0.14160771667957306,
|
| 645 |
+
"learning_rate": 1.9725819080487976e-05,
|
| 646 |
+
"loss": 0.1749,
|
| 647 |
+
"step": 780
|
| 648 |
+
},
|
| 649 |
+
{
|
| 650 |
+
"epoch": 0.09665973326807782,
|
| 651 |
+
"grad_norm": 0.17038081586360931,
|
| 652 |
+
"learning_rate": 1.970647517130819e-05,
|
| 653 |
+
"loss": 0.1795,
|
| 654 |
+
"step": 790
|
| 655 |
+
},
|
| 656 |
+
{
|
| 657 |
+
"epoch": 0.09788327419552184,
|
| 658 |
+
"grad_norm": 0.24339532852172852,
|
| 659 |
+
"learning_rate": 1.968648209398759e-05,
|
| 660 |
+
"loss": 0.161,
|
| 661 |
+
"step": 800
|
| 662 |
+
},
|
| 663 |
+
{
|
| 664 |
+
"epoch": 0.09788327419552184,
|
| 665 |
+
"eval_accuracy": 0.95721,
|
| 666 |
+
"eval_f1_weighted": 0.9572295598206603,
|
| 667 |
+
"eval_loss": 0.03337462991476059,
|
| 668 |
+
"eval_macro_f1": 0.9571669846830866,
|
| 669 |
+
"eval_precision": 0.9576467364395861,
|
| 670 |
+
"eval_recall": 0.95721,
|
| 671 |
+
"eval_runtime": 759.9975,
|
| 672 |
+
"eval_samples_per_second": 131.579,
|
| 673 |
+
"eval_steps_per_second": 1.029,
|
| 674 |
+
"step": 800
|
| 675 |
+
},
|
| 676 |
+
{
|
| 677 |
+
"epoch": 0.09910681512296586,
|
| 678 |
+
"grad_norm": 0.16346058249473572,
|
| 679 |
+
"learning_rate": 1.9665841185661283e-05,
|
| 680 |
+
"loss": 0.1596,
|
| 681 |
+
"step": 810
|
| 682 |
+
},
|
| 683 |
+
{
|
| 684 |
+
"epoch": 0.10033035605040988,
|
| 685 |
+
"grad_norm": 0.25031423568725586,
|
| 686 |
+
"learning_rate": 1.964455382679128e-05,
|
| 687 |
+
"loss": 0.1662,
|
| 688 |
+
"step": 820
|
| 689 |
+
},
|
| 690 |
+
{
|
| 691 |
+
"epoch": 0.10155389697785391,
|
| 692 |
+
"grad_norm": 0.36678382754325867,
|
| 693 |
+
"learning_rate": 1.9622621441074113e-05,
|
| 694 |
+
"loss": 0.1552,
|
| 695 |
+
"step": 830
|
| 696 |
+
},
|
| 697 |
+
{
|
| 698 |
+
"epoch": 0.10277743790529793,
|
| 699 |
+
"grad_norm": 0.34810611605644226,
|
| 700 |
+
"learning_rate": 1.960004549534567e-05,
|
| 701 |
+
"loss": 0.1663,
|
| 702 |
+
"step": 840
|
| 703 |
+
},
|
| 704 |
+
{
|
| 705 |
+
"epoch": 0.10400097883274195,
|
| 706 |
+
"grad_norm": 0.15511688590049744,
|
| 707 |
+
"learning_rate": 1.957682749948307e-05,
|
| 708 |
+
"loss": 0.1571,
|
| 709 |
+
"step": 850
|
| 710 |
+
},
|
| 711 |
+
{
|
| 712 |
+
"epoch": 0.10522451976018597,
|
| 713 |
+
"grad_norm": 0.2082190364599228,
|
| 714 |
+
"learning_rate": 1.9552969006303666e-05,
|
| 715 |
+
"loss": 0.1497,
|
| 716 |
+
"step": 860
|
| 717 |
+
},
|
| 718 |
+
{
|
| 719 |
+
"epoch": 0.10644806068763,
|
| 720 |
+
"grad_norm": 0.14936988055706024,
|
| 721 |
+
"learning_rate": 1.9528471611461235e-05,
|
| 722 |
+
"loss": 0.1435,
|
| 723 |
+
"step": 870
|
| 724 |
+
},
|
| 725 |
+
{
|
| 726 |
+
"epoch": 0.10767160161507403,
|
| 727 |
+
"grad_norm": 0.252530574798584,
|
| 728 |
+
"learning_rate": 1.9503336953339223e-05,
|
| 729 |
+
"loss": 0.1536,
|
| 730 |
+
"step": 880
|
| 731 |
+
},
|
| 732 |
+
{
|
| 733 |
+
"epoch": 0.10889514254251804,
|
| 734 |
+
"grad_norm": 0.07253506034612656,
|
| 735 |
+
"learning_rate": 1.947756671294118e-05,
|
| 736 |
+
"loss": 0.1605,
|
| 737 |
+
"step": 890
|
| 738 |
+
},
|
| 739 |
+
{
|
| 740 |
+
"epoch": 0.11011868346996206,
|
| 741 |
+
"grad_norm": 0.09060605615377426,
|
| 742 |
+
"learning_rate": 1.9451162613778353e-05,
|
| 743 |
+
"loss": 0.1581,
|
| 744 |
+
"step": 900
|
| 745 |
+
},
|
| 746 |
+
{
|
| 747 |
+
"epoch": 0.11011868346996206,
|
| 748 |
+
"eval_accuracy": 0.95459,
|
| 749 |
+
"eval_f1_weighted": 0.9546135918699761,
|
| 750 |
+
"eval_loss": 0.03526490181684494,
|
| 751 |
+
"eval_macro_f1": 0.9545725664001194,
|
| 752 |
+
"eval_precision": 0.9559789658562307,
|
| 753 |
+
"eval_recall": 0.95459,
|
| 754 |
+
"eval_runtime": 759.7202,
|
| 755 |
+
"eval_samples_per_second": 131.627,
|
| 756 |
+
"eval_steps_per_second": 1.029,
|
| 757 |
+
"step": 900
|
| 758 |
+
},
|
| 759 |
+
{
|
| 760 |
+
"epoch": 0.1113422243974061,
|
| 761 |
+
"grad_norm": 0.21080073714256287,
|
| 762 |
+
"learning_rate": 1.942412642175439e-05,
|
| 763 |
+
"loss": 0.164,
|
| 764 |
+
"step": 910
|
| 765 |
+
},
|
| 766 |
+
{
|
| 767 |
+
"epoch": 0.11256576532485012,
|
| 768 |
+
"grad_norm": 0.22832269966602325,
|
| 769 |
+
"learning_rate": 1.939645994504726e-05,
|
| 770 |
+
"loss": 0.1595,
|
| 771 |
+
"step": 920
|
| 772 |
+
},
|
| 773 |
+
{
|
| 774 |
+
"epoch": 0.11378930625229414,
|
| 775 |
+
"grad_norm": 0.19044892489910126,
|
| 776 |
+
"learning_rate": 1.93681650339883e-05,
|
| 777 |
+
"loss": 0.1648,
|
| 778 |
+
"step": 930
|
| 779 |
+
},
|
| 780 |
+
{
|
| 781 |
+
"epoch": 0.11501284717973816,
|
| 782 |
+
"grad_norm": 0.2413518726825714,
|
| 783 |
+
"learning_rate": 1.9339243580938493e-05,
|
| 784 |
+
"loss": 0.155,
|
| 785 |
+
"step": 940
|
| 786 |
+
},
|
| 787 |
+
{
|
| 788 |
+
"epoch": 0.11623638810718219,
|
| 789 |
+
"grad_norm": 0.13218896090984344,
|
| 790 |
+
"learning_rate": 1.930969752016187e-05,
|
| 791 |
+
"loss": 0.1499,
|
| 792 |
+
"step": 950
|
| 793 |
+
},
|
| 794 |
+
{
|
| 795 |
+
"epoch": 0.11745992903462621,
|
| 796 |
+
"grad_norm": 0.23770441114902496,
|
| 797 |
+
"learning_rate": 1.927952882769619e-05,
|
| 798 |
+
"loss": 0.1555,
|
| 799 |
+
"step": 960
|
| 800 |
+
},
|
| 801 |
+
{
|
| 802 |
+
"epoch": 0.11868346996207023,
|
| 803 |
+
"grad_norm": 0.2193710207939148,
|
| 804 |
+
"learning_rate": 1.924873952122074e-05,
|
| 805 |
+
"loss": 0.1505,
|
| 806 |
+
"step": 970
|
| 807 |
+
},
|
| 808 |
+
{
|
| 809 |
+
"epoch": 0.11990701088951425,
|
| 810 |
+
"grad_norm": 0.1577174812555313,
|
| 811 |
+
"learning_rate": 1.9217331659921432e-05,
|
| 812 |
+
"loss": 0.1551,
|
| 813 |
+
"step": 980
|
| 814 |
+
},
|
| 815 |
+
{
|
| 816 |
+
"epoch": 0.12113055181695828,
|
| 817 |
+
"grad_norm": 0.1814967393875122,
|
| 818 |
+
"learning_rate": 1.9185307344353052e-05,
|
| 819 |
+
"loss": 0.1607,
|
| 820 |
+
"step": 990
|
| 821 |
+
},
|
| 822 |
+
{
|
| 823 |
+
"epoch": 0.1223540927444023,
|
| 824 |
+
"grad_norm": 0.27239707112312317,
|
| 825 |
+
"learning_rate": 1.91526687162988e-05,
|
| 826 |
+
"loss": 0.1506,
|
| 827 |
+
"step": 1000
|
| 828 |
+
},
|
| 829 |
+
{
|
| 830 |
+
"epoch": 0.1223540927444023,
|
| 831 |
+
"eval_accuracy": 0.96454,
|
| 832 |
+
"eval_f1_weighted": 0.9645510150201642,
|
| 833 |
+
"eval_loss": 0.028770776465535164,
|
| 834 |
+
"eval_macro_f1": 0.9644889377060355,
|
| 835 |
+
"eval_precision": 0.9646912114769961,
|
| 836 |
+
"eval_recall": 0.96454,
|
| 837 |
+
"eval_runtime": 759.5962,
|
| 838 |
+
"eval_samples_per_second": 131.649,
|
| 839 |
+
"eval_steps_per_second": 1.029,
|
| 840 |
+
"step": 1000
|
| 841 |
+
},
|
| 842 |
+
{
|
| 843 |
+
"epoch": 0.12357763367184632,
|
| 844 |
+
"grad_norm": 0.20511822402477264,
|
| 845 |
+
"learning_rate": 1.9119417958627022e-05,
|
| 846 |
+
"loss": 0.1503,
|
| 847 |
+
"step": 1010
|
| 848 |
+
},
|
| 849 |
+
{
|
| 850 |
+
"epoch": 0.12480117459929034,
|
| 851 |
+
"grad_norm": 0.2324151247739792,
|
| 852 |
+
"learning_rate": 1.9085557295145254e-05,
|
| 853 |
+
"loss": 0.1581,
|
| 854 |
+
"step": 1020
|
| 855 |
+
},
|
| 856 |
+
{
|
| 857 |
+
"epoch": 0.12602471552673436,
|
| 858 |
+
"grad_norm": 0.11019974201917648,
|
| 859 |
+
"learning_rate": 1.905108899045146e-05,
|
| 860 |
+
"loss": 0.1556,
|
| 861 |
+
"step": 1030
|
| 862 |
+
},
|
| 863 |
+
{
|
| 864 |
+
"epoch": 0.1272482564541784,
|
| 865 |
+
"grad_norm": 0.13370558619499207,
|
| 866 |
+
"learning_rate": 1.9016015349782594e-05,
|
| 867 |
+
"loss": 0.1469,
|
| 868 |
+
"step": 1040
|
| 869 |
+
},
|
| 870 |
+
{
|
| 871 |
+
"epoch": 0.12847179738162243,
|
| 872 |
+
"grad_norm": 0.11238931119441986,
|
| 873 |
+
"learning_rate": 1.8980338718860426e-05,
|
| 874 |
+
"loss": 0.155,
|
| 875 |
+
"step": 1050
|
| 876 |
+
},
|
| 877 |
+
{
|
| 878 |
+
"epoch": 0.12969533830906643,
|
| 879 |
+
"grad_norm": 0.2959650158882141,
|
| 880 |
+
"learning_rate": 1.894406148373466e-05,
|
| 881 |
+
"loss": 0.1478,
|
| 882 |
+
"step": 1060
|
| 883 |
+
},
|
| 884 |
+
{
|
| 885 |
+
"epoch": 0.13091887923651047,
|
| 886 |
+
"grad_norm": 0.14401015639305115,
|
| 887 |
+
"learning_rate": 1.890718607062334e-05,
|
| 888 |
+
"loss": 0.1558,
|
| 889 |
+
"step": 1070
|
| 890 |
+
},
|
| 891 |
+
{
|
| 892 |
+
"epoch": 0.13214242016395447,
|
| 893 |
+
"grad_norm": 0.2187667042016983,
|
| 894 |
+
"learning_rate": 1.8869714945750614e-05,
|
| 895 |
+
"loss": 0.1618,
|
| 896 |
+
"step": 1080
|
| 897 |
+
},
|
| 898 |
+
{
|
| 899 |
+
"epoch": 0.1333659610913985,
|
| 900 |
+
"grad_norm": 0.17045481503009796,
|
| 901 |
+
"learning_rate": 1.883165061518176e-05,
|
| 902 |
+
"loss": 0.1487,
|
| 903 |
+
"step": 1090
|
| 904 |
+
},
|
| 905 |
+
{
|
| 906 |
+
"epoch": 0.13458950201884254,
|
| 907 |
+
"grad_norm": 0.09508351981639862,
|
| 908 |
+
"learning_rate": 1.8792995624655605e-05,
|
| 909 |
+
"loss": 0.1461,
|
| 910 |
+
"step": 1100
|
| 911 |
+
},
|
| 912 |
+
{
|
| 913 |
+
"epoch": 0.13458950201884254,
|
| 914 |
+
"eval_accuracy": 0.95198,
|
| 915 |
+
"eval_f1_weighted": 0.9520023284947491,
|
| 916 |
+
"eval_loss": 0.04088817164301872,
|
| 917 |
+
"eval_macro_f1": 0.9519693367124435,
|
| 918 |
+
"eval_precision": 0.9538785423213092,
|
| 919 |
+
"eval_recall": 0.95198,
|
| 920 |
+
"eval_runtime": 759.8551,
|
| 921 |
+
"eval_samples_per_second": 131.604,
|
| 922 |
+
"eval_steps_per_second": 1.029,
|
| 923 |
+
"step": 1100
|
| 924 |
+
},
|
| 925 |
+
{
|
| 926 |
+
"epoch": 0.13581304294628654,
|
| 927 |
+
"grad_norm": 0.1880544275045395,
|
| 928 |
+
"learning_rate": 1.875375255941426e-05,
|
| 929 |
+
"loss": 0.1442,
|
| 930 |
+
"step": 1110
|
| 931 |
+
},
|
| 932 |
+
{
|
| 933 |
+
"epoch": 0.13703658387373058,
|
| 934 |
+
"grad_norm": 0.1614997684955597,
|
| 935 |
+
"learning_rate": 1.8713924044030212e-05,
|
| 936 |
+
"loss": 0.1622,
|
| 937 |
+
"step": 1120
|
| 938 |
+
},
|
| 939 |
+
{
|
| 940 |
+
"epoch": 0.1382601248011746,
|
| 941 |
+
"grad_norm": 0.10545000433921814,
|
| 942 |
+
"learning_rate": 1.86735127422308e-05,
|
| 943 |
+
"loss": 0.1531,
|
| 944 |
+
"step": 1130
|
| 945 |
+
},
|
| 946 |
+
{
|
| 947 |
+
"epoch": 0.13948366572861862,
|
| 948 |
+
"grad_norm": 0.2203921526670456,
|
| 949 |
+
"learning_rate": 1.8632521356720064e-05,
|
| 950 |
+
"loss": 0.1471,
|
| 951 |
+
"step": 1140
|
| 952 |
+
},
|
| 953 |
+
{
|
| 954 |
+
"epoch": 0.14070720665606265,
|
| 955 |
+
"grad_norm": 0.16150350868701935,
|
| 956 |
+
"learning_rate": 1.8590952628997994e-05,
|
| 957 |
+
"loss": 0.1436,
|
| 958 |
+
"step": 1150
|
| 959 |
+
},
|
| 960 |
+
{
|
| 961 |
+
"epoch": 0.14193074758350666,
|
| 962 |
+
"grad_norm": 0.1466459184885025,
|
| 963 |
+
"learning_rate": 1.8548809339177155e-05,
|
| 964 |
+
"loss": 0.1486,
|
| 965 |
+
"step": 1160
|
| 966 |
+
},
|
| 967 |
+
{
|
| 968 |
+
"epoch": 0.1431542885109507,
|
| 969 |
+
"grad_norm": 0.1858566850423813,
|
| 970 |
+
"learning_rate": 1.8506094305796786e-05,
|
| 971 |
+
"loss": 0.1485,
|
| 972 |
+
"step": 1170
|
| 973 |
+
},
|
| 974 |
+
{
|
| 975 |
+
"epoch": 0.14437782943839472,
|
| 976 |
+
"grad_norm": 0.21629975736141205,
|
| 977 |
+
"learning_rate": 1.846281038563428e-05,
|
| 978 |
+
"loss": 0.1558,
|
| 979 |
+
"step": 1180
|
| 980 |
+
},
|
| 981 |
+
{
|
| 982 |
+
"epoch": 0.14560137036583873,
|
| 983 |
+
"grad_norm": 0.12939126789569855,
|
| 984 |
+
"learning_rate": 1.841896047351412e-05,
|
| 985 |
+
"loss": 0.1388,
|
| 986 |
+
"step": 1190
|
| 987 |
+
},
|
| 988 |
+
{
|
| 989 |
+
"epoch": 0.14682491129328276,
|
| 990 |
+
"grad_norm": 0.12447889149188995,
|
| 991 |
+
"learning_rate": 1.8374547502114277e-05,
|
| 992 |
+
"loss": 0.1683,
|
| 993 |
+
"step": 1200
|
| 994 |
+
},
|
| 995 |
+
{
|
| 996 |
+
"epoch": 0.14682491129328276,
|
| 997 |
+
"eval_accuracy": 0.96627,
|
| 998 |
+
"eval_f1_weighted": 0.9662808443914441,
|
| 999 |
+
"eval_loss": 0.027206970378756523,
|
| 1000 |
+
"eval_macro_f1": 0.9662223734621376,
|
| 1001 |
+
"eval_precision": 0.966433068847901,
|
| 1002 |
+
"eval_recall": 0.96627,
|
| 1003 |
+
"eval_runtime": 661.8465,
|
| 1004 |
+
"eval_samples_per_second": 151.092,
|
| 1005 |
+
"eval_steps_per_second": 1.182,
|
| 1006 |
+
"step": 1200
|
| 1007 |
+
},
|
| 1008 |
+
{
|
| 1009 |
+
"epoch": 0.1480484522207268,
|
| 1010 |
+
"grad_norm": 0.07575482875108719,
|
| 1011 |
+
"learning_rate": 1.832957444177007e-05,
|
| 1012 |
+
"loss": 0.152,
|
| 1013 |
+
"step": 1210
|
| 1014 |
+
},
|
| 1015 |
+
{
|
| 1016 |
+
"epoch": 0.1492719931481708,
|
| 1017 |
+
"grad_norm": 0.22879591584205627,
|
| 1018 |
+
"learning_rate": 1.828404430027553e-05,
|
| 1019 |
+
"loss": 0.15,
|
| 1020 |
+
"step": 1220
|
| 1021 |
+
},
|
| 1022 |
+
{
|
| 1023 |
+
"epoch": 0.15049553407561483,
|
| 1024 |
+
"grad_norm": 0.25837475061416626,
|
| 1025 |
+
"learning_rate": 1.8237960122682207e-05,
|
| 1026 |
+
"loss": 0.1557,
|
| 1027 |
+
"step": 1230
|
| 1028 |
+
},
|
| 1029 |
+
{
|
| 1030 |
+
"epoch": 0.15171907500305884,
|
| 1031 |
+
"grad_norm": 0.1440538763999939,
|
| 1032 |
+
"learning_rate": 1.8191324991095527e-05,
|
| 1033 |
+
"loss": 0.154,
|
| 1034 |
+
"step": 1240
|
| 1035 |
+
},
|
| 1036 |
+
{
|
| 1037 |
+
"epoch": 0.15294261593050287,
|
| 1038 |
+
"grad_norm": 0.21304763853549957,
|
| 1039 |
+
"learning_rate": 1.814414202446869e-05,
|
| 1040 |
+
"loss": 0.1428,
|
| 1041 |
+
"step": 1250
|
| 1042 |
+
},
|
| 1043 |
+
{
|
| 1044 |
+
"epoch": 0.1541661568579469,
|
| 1045 |
+
"grad_norm": 0.2292693555355072,
|
| 1046 |
+
"learning_rate": 1.809641437839403e-05,
|
| 1047 |
+
"loss": 0.1441,
|
| 1048 |
+
"step": 1260
|
| 1049 |
+
},
|
| 1050 |
+
{
|
| 1051 |
+
"epoch": 0.1553896977853909,
|
| 1052 |
+
"grad_norm": 0.0891769602894783,
|
| 1053 |
+
"learning_rate": 1.8048145244892e-05,
|
| 1054 |
+
"loss": 0.1563,
|
| 1055 |
+
"step": 1270
|
| 1056 |
+
},
|
| 1057 |
+
{
|
| 1058 |
+
"epoch": 0.15661323871283495,
|
| 1059 |
+
"grad_norm": 0.13292840123176575,
|
| 1060 |
+
"learning_rate": 1.799933785219769e-05,
|
| 1061 |
+
"loss": 0.1418,
|
| 1062 |
+
"step": 1280
|
| 1063 |
+
},
|
| 1064 |
+
{
|
| 1065 |
+
"epoch": 0.15783677964027898,
|
| 1066 |
+
"grad_norm": 0.30031099915504456,
|
| 1067 |
+
"learning_rate": 1.7949995464544892e-05,
|
| 1068 |
+
"loss": 0.1543,
|
| 1069 |
+
"step": 1290
|
| 1070 |
+
},
|
| 1071 |
+
{
|
| 1072 |
+
"epoch": 0.15906032056772298,
|
| 1073 |
+
"grad_norm": 0.08812980353832245,
|
| 1074 |
+
"learning_rate": 1.790012138194783e-05,
|
| 1075 |
+
"loss": 0.1473,
|
| 1076 |
+
"step": 1300
|
| 1077 |
+
},
|
| 1078 |
+
{
|
| 1079 |
+
"epoch": 0.15906032056772298,
|
| 1080 |
+
"eval_accuracy": 0.96636,
|
| 1081 |
+
"eval_f1_weighted": 0.9663763998373325,
|
| 1082 |
+
"eval_loss": 0.02714269980788231,
|
| 1083 |
+
"eval_macro_f1": 0.9663303002945869,
|
| 1084 |
+
"eval_precision": 0.9669126928794434,
|
| 1085 |
+
"eval_recall": 0.96636,
|
| 1086 |
+
"eval_runtime": 661.9303,
|
| 1087 |
+
"eval_samples_per_second": 151.073,
|
| 1088 |
+
"eval_steps_per_second": 1.181,
|
| 1089 |
+
"step": 1300
|
| 1090 |
+
},
|
| 1091 |
+
{
|
| 1092 |
+
"epoch": 0.16028386149516702,
|
| 1093 |
+
"grad_norm": 0.1380453258752823,
|
| 1094 |
+
"learning_rate": 1.7849718939980426e-05,
|
| 1095 |
+
"loss": 0.1594,
|
| 1096 |
+
"step": 1310
|
| 1097 |
+
},
|
| 1098 |
+
{
|
| 1099 |
+
"epoch": 0.16150740242261102,
|
| 1100 |
+
"grad_norm": 0.08388533443212509,
|
| 1101 |
+
"learning_rate": 1.7798791509553224e-05,
|
| 1102 |
+
"loss": 0.1411,
|
| 1103 |
+
"step": 1320
|
| 1104 |
+
},
|
| 1105 |
+
{
|
| 1106 |
+
"epoch": 0.16273094335005506,
|
| 1107 |
+
"grad_norm": 0.2659275531768799,
|
| 1108 |
+
"learning_rate": 1.7747342496687957e-05,
|
| 1109 |
+
"loss": 0.1495,
|
| 1110 |
+
"step": 1330
|
| 1111 |
+
},
|
| 1112 |
+
{
|
| 1113 |
+
"epoch": 0.1639544842774991,
|
| 1114 |
+
"grad_norm": 0.275398313999176,
|
| 1115 |
+
"learning_rate": 1.7695375342289732e-05,
|
| 1116 |
+
"loss": 0.134,
|
| 1117 |
+
"step": 1340
|
| 1118 |
+
},
|
| 1119 |
+
{
|
| 1120 |
+
"epoch": 0.1651780252049431,
|
| 1121 |
+
"grad_norm": 0.15272289514541626,
|
| 1122 |
+
"learning_rate": 1.7642893521916926e-05,
|
| 1123 |
+
"loss": 0.1427,
|
| 1124 |
+
"step": 1350
|
| 1125 |
+
},
|
| 1126 |
+
{
|
| 1127 |
+
"epoch": 0.16640156613238713,
|
| 1128 |
+
"grad_norm": 0.21478359401226044,
|
| 1129 |
+
"learning_rate": 1.7589900545548706e-05,
|
| 1130 |
+
"loss": 0.1465,
|
| 1131 |
+
"step": 1360
|
| 1132 |
+
},
|
| 1133 |
+
{
|
| 1134 |
+
"epoch": 0.16762510705983116,
|
| 1135 |
+
"grad_norm": 0.2184278517961502,
|
| 1136 |
+
"learning_rate": 1.7536399957350326e-05,
|
| 1137 |
+
"loss": 0.1452,
|
| 1138 |
+
"step": 1370
|
| 1139 |
+
},
|
| 1140 |
+
{
|
| 1141 |
+
"epoch": 0.16884864798727517,
|
| 1142 |
+
"grad_norm": 0.23272652924060822,
|
| 1143 |
+
"learning_rate": 1.7482395335436052e-05,
|
| 1144 |
+
"loss": 0.142,
|
| 1145 |
+
"step": 1380
|
| 1146 |
+
},
|
| 1147 |
+
{
|
| 1148 |
+
"epoch": 0.1700721889147192,
|
| 1149 |
+
"grad_norm": 0.14892522990703583,
|
| 1150 |
+
"learning_rate": 1.7427890291629895e-05,
|
| 1151 |
+
"loss": 0.1465,
|
| 1152 |
+
"step": 1390
|
| 1153 |
+
},
|
| 1154 |
+
{
|
| 1155 |
+
"epoch": 0.1712957298421632,
|
| 1156 |
+
"grad_norm": 0.2066282033920288,
|
| 1157 |
+
"learning_rate": 1.737288847122402e-05,
|
| 1158 |
+
"loss": 0.1444,
|
| 1159 |
+
"step": 1400
|
| 1160 |
+
},
|
| 1161 |
+
{
|
| 1162 |
+
"epoch": 0.1712957298421632,
|
| 1163 |
+
"eval_accuracy": 0.95302,
|
| 1164 |
+
"eval_f1_weighted": 0.9530351521706546,
|
| 1165 |
+
"eval_loss": 0.03864819556474686,
|
| 1166 |
+
"eval_macro_f1": 0.9530165251021966,
|
| 1167 |
+
"eval_precision": 0.9557329367218795,
|
| 1168 |
+
"eval_recall": 0.95302,
|
| 1169 |
+
"eval_runtime": 661.7384,
|
| 1170 |
+
"eval_samples_per_second": 151.117,
|
| 1171 |
+
"eval_steps_per_second": 1.182,
|
| 1172 |
+
"step": 1400
|
| 1173 |
+
},
|
| 1174 |
+
{
|
| 1175 |
+
"epoch": 0.17251927076960724,
|
| 1176 |
+
"grad_norm": 0.2941270172595978,
|
| 1177 |
+
"learning_rate": 1.7317393552734977e-05,
|
| 1178 |
+
"loss": 0.136,
|
| 1179 |
+
"step": 1410
|
| 1180 |
+
},
|
| 1181 |
+
{
|
| 1182 |
+
"epoch": 0.17374281169705127,
|
| 1183 |
+
"grad_norm": 0.11988822370767593,
|
| 1184 |
+
"learning_rate": 1.7261409247657656e-05,
|
| 1185 |
+
"loss": 0.1551,
|
| 1186 |
+
"step": 1420
|
| 1187 |
+
},
|
| 1188 |
+
{
|
| 1189 |
+
"epoch": 0.17496635262449528,
|
| 1190 |
+
"grad_norm": 0.346236914396286,
|
| 1191 |
+
"learning_rate": 1.7204939300217084e-05,
|
| 1192 |
+
"loss": 0.1453,
|
| 1193 |
+
"step": 1430
|
| 1194 |
+
},
|
| 1195 |
+
{
|
| 1196 |
+
"epoch": 0.17618989355193931,
|
| 1197 |
+
"grad_norm": 0.15973429381847382,
|
| 1198 |
+
"learning_rate": 1.7147987487118006e-05,
|
| 1199 |
+
"loss": 0.1341,
|
| 1200 |
+
"step": 1440
|
| 1201 |
+
},
|
| 1202 |
+
{
|
| 1203 |
+
"epoch": 0.17741343447938335,
|
| 1204 |
+
"grad_norm": 0.2531983554363251,
|
| 1205 |
+
"learning_rate": 1.7090557617292287e-05,
|
| 1206 |
+
"loss": 0.1479,
|
| 1207 |
+
"step": 1450
|
| 1208 |
+
},
|
| 1209 |
+
{
|
| 1210 |
+
"epoch": 0.17863697540682735,
|
| 1211 |
+
"grad_norm": 0.10105066746473312,
|
| 1212 |
+
"learning_rate": 1.703265353164419e-05,
|
| 1213 |
+
"loss": 0.1459,
|
| 1214 |
+
"step": 1460
|
| 1215 |
+
},
|
| 1216 |
+
{
|
| 1217 |
+
"epoch": 0.1798605163342714,
|
| 1218 |
+
"grad_norm": 0.10651528090238571,
|
| 1219 |
+
"learning_rate": 1.6974279102793473e-05,
|
| 1220 |
+
"loss": 0.152,
|
| 1221 |
+
"step": 1470
|
| 1222 |
+
},
|
| 1223 |
+
{
|
| 1224 |
+
"epoch": 0.1810840572617154,
|
| 1225 |
+
"grad_norm": 0.2675159275531769,
|
| 1226 |
+
"learning_rate": 1.6915438234816423e-05,
|
| 1227 |
+
"loss": 0.1467,
|
| 1228 |
+
"step": 1480
|
| 1229 |
+
},
|
| 1230 |
+
{
|
| 1231 |
+
"epoch": 0.18230759818915943,
|
| 1232 |
+
"grad_norm": 0.09507659822702408,
|
| 1233 |
+
"learning_rate": 1.6856134862984718e-05,
|
| 1234 |
+
"loss": 0.1504,
|
| 1235 |
+
"step": 1490
|
| 1236 |
+
},
|
| 1237 |
+
{
|
| 1238 |
+
"epoch": 0.18353113911660346,
|
| 1239 |
+
"grad_norm": 0.13831426203250885,
|
| 1240 |
+
"learning_rate": 1.6796372953502252e-05,
|
| 1241 |
+
"loss": 0.151,
|
| 1242 |
+
"step": 1500
|
| 1243 |
+
},
|
| 1244 |
+
{
|
| 1245 |
+
"epoch": 0.18353113911660346,
|
| 1246 |
+
"eval_accuracy": 0.96716,
|
| 1247 |
+
"eval_f1_weighted": 0.9671757267424962,
|
| 1248 |
+
"eval_loss": 0.02643330581486225,
|
| 1249 |
+
"eval_macro_f1": 0.9671297823300075,
|
| 1250 |
+
"eval_precision": 0.9676731949933907,
|
| 1251 |
+
"eval_recall": 0.96716,
|
| 1252 |
+
"eval_runtime": 662.2285,
|
| 1253 |
+
"eval_samples_per_second": 151.005,
|
| 1254 |
+
"eval_steps_per_second": 1.181,
|
| 1255 |
+
"step": 1500
|
| 1256 |
+
},
|
| 1257 |
+
{
|
| 1258 |
+
"epoch": 0.18475468004404746,
|
| 1259 |
+
"grad_norm": 0.20989136397838593,
|
| 1260 |
+
"learning_rate": 1.673615650323988e-05,
|
| 1261 |
+
"loss": 0.1424,
|
| 1262 |
+
"step": 1510
|
| 1263 |
+
},
|
| 1264 |
+
{
|
| 1265 |
+
"epoch": 0.1859782209714915,
|
| 1266 |
+
"grad_norm": 0.13308484852313995,
|
| 1267 |
+
"learning_rate": 1.6675489539468094e-05,
|
| 1268 |
+
"loss": 0.1541,
|
| 1269 |
+
"step": 1520
|
| 1270 |
+
},
|
| 1271 |
+
{
|
| 1272 |
+
"epoch": 0.18720176189893553,
|
| 1273 |
+
"grad_norm": 0.3802664577960968,
|
| 1274 |
+
"learning_rate": 1.6614376119587685e-05,
|
| 1275 |
+
"loss": 0.1426,
|
| 1276 |
+
"step": 1530
|
| 1277 |
+
},
|
| 1278 |
+
{
|
| 1279 |
+
"epoch": 0.18842530282637954,
|
| 1280 |
+
"grad_norm": 0.254869282245636,
|
| 1281 |
+
"learning_rate": 1.6552820330858398e-05,
|
| 1282 |
+
"loss": 0.1407,
|
| 1283 |
+
"step": 1540
|
| 1284 |
+
},
|
| 1285 |
+
{
|
| 1286 |
+
"epoch": 0.18964884375382357,
|
| 1287 |
+
"grad_norm": 0.22702300548553467,
|
| 1288 |
+
"learning_rate": 1.6490826290125555e-05,
|
| 1289 |
+
"loss": 0.1404,
|
| 1290 |
+
"step": 1550
|
| 1291 |
+
},
|
| 1292 |
+
{
|
| 1293 |
+
"epoch": 0.19087238468126758,
|
| 1294 |
+
"grad_norm": 0.18147043883800507,
|
| 1295 |
+
"learning_rate": 1.6428398143544727e-05,
|
| 1296 |
+
"loss": 0.1398,
|
| 1297 |
+
"step": 1560
|
| 1298 |
+
},
|
| 1299 |
+
{
|
| 1300 |
+
"epoch": 0.1920959256087116,
|
| 1301 |
+
"grad_norm": 0.2697111964225769,
|
| 1302 |
+
"learning_rate": 1.636554006630446e-05,
|
| 1303 |
+
"loss": 0.1465,
|
| 1304 |
+
"step": 1570
|
| 1305 |
+
},
|
| 1306 |
+
{
|
| 1307 |
+
"epoch": 0.19331946653615564,
|
| 1308 |
+
"grad_norm": 0.22285781800746918,
|
| 1309 |
+
"learning_rate": 1.6302256262347e-05,
|
| 1310 |
+
"loss": 0.1398,
|
| 1311 |
+
"step": 1580
|
| 1312 |
+
},
|
| 1313 |
+
{
|
| 1314 |
+
"epoch": 0.19454300746359965,
|
| 1315 |
+
"grad_norm": 0.2551407814025879,
|
| 1316 |
+
"learning_rate": 1.6238550964087175e-05,
|
| 1317 |
+
"loss": 0.1347,
|
| 1318 |
+
"step": 1590
|
| 1319 |
+
},
|
| 1320 |
+
{
|
| 1321 |
+
"epoch": 0.19576654839104368,
|
| 1322 |
+
"grad_norm": 0.25052276253700256,
|
| 1323 |
+
"learning_rate": 1.61744284321293e-05,
|
| 1324 |
+
"loss": 0.1461,
|
| 1325 |
+
"step": 1600
|
| 1326 |
+
},
|
| 1327 |
+
{
|
| 1328 |
+
"epoch": 0.19576654839104368,
|
| 1329 |
+
"eval_accuracy": 0.96526,
|
| 1330 |
+
"eval_f1_weighted": 0.9652784309636304,
|
| 1331 |
+
"eval_loss": 0.02899531088769436,
|
| 1332 |
+
"eval_macro_f1": 0.9652431164406312,
|
| 1333 |
+
"eval_precision": 0.9664143950990177,
|
| 1334 |
+
"eval_recall": 0.96526,
|
| 1335 |
+
"eval_runtime": 662.3587,
|
| 1336 |
+
"eval_samples_per_second": 150.976,
|
| 1337 |
+
"eval_steps_per_second": 1.181,
|
| 1338 |
+
"step": 1600
|
| 1339 |
+
},
|
| 1340 |
+
{
|
| 1341 |
+
"epoch": 0.19699008931848772,
|
| 1342 |
+
"grad_norm": 0.09207922220230103,
|
| 1343 |
+
"learning_rate": 1.6109892954982242e-05,
|
| 1344 |
+
"loss": 0.1382,
|
| 1345 |
+
"step": 1610
|
| 1346 |
+
},
|
| 1347 |
+
{
|
| 1348 |
+
"epoch": 0.19821363024593172,
|
| 1349 |
+
"grad_norm": 0.2670917809009552,
|
| 1350 |
+
"learning_rate": 1.6044948848772605e-05,
|
| 1351 |
+
"loss": 0.142,
|
| 1352 |
+
"step": 1620
|
| 1353 |
+
},
|
| 1354 |
+
{
|
| 1355 |
+
"epoch": 0.19943717117337575,
|
| 1356 |
+
"grad_norm": 0.28024110198020935,
|
| 1357 |
+
"learning_rate": 1.597960045695607e-05,
|
| 1358 |
+
"loss": 0.1385,
|
| 1359 |
+
"step": 1630
|
| 1360 |
+
},
|
| 1361 |
+
{
|
| 1362 |
+
"epoch": 0.20066071210081976,
|
| 1363 |
+
"grad_norm": 0.15802879631519318,
|
| 1364 |
+
"learning_rate": 1.5913852150026895e-05,
|
| 1365 |
+
"loss": 0.1328,
|
| 1366 |
+
"step": 1640
|
| 1367 |
+
},
|
| 1368 |
+
{
|
| 1369 |
+
"epoch": 0.2018842530282638,
|
| 1370 |
+
"grad_norm": 0.12178096175193787,
|
| 1371 |
+
"learning_rate": 1.5847708325225618e-05,
|
| 1372 |
+
"loss": 0.1446,
|
| 1373 |
+
"step": 1650
|
| 1374 |
+
},
|
| 1375 |
+
{
|
| 1376 |
+
"epoch": 0.20310779395570783,
|
| 1377 |
+
"grad_norm": 0.22587499022483826,
|
| 1378 |
+
"learning_rate": 1.578117340624499e-05,
|
| 1379 |
+
"loss": 0.1406,
|
| 1380 |
+
"step": 1660
|
| 1381 |
+
},
|
| 1382 |
+
{
|
| 1383 |
+
"epoch": 0.20433133488315183,
|
| 1384 |
+
"grad_norm": 0.18149550259113312,
|
| 1385 |
+
"learning_rate": 1.5714251842934096e-05,
|
| 1386 |
+
"loss": 0.1459,
|
| 1387 |
+
"step": 1670
|
| 1388 |
+
},
|
| 1389 |
+
{
|
| 1390 |
+
"epoch": 0.20555487581059587,
|
| 1391 |
+
"grad_norm": 0.15662655234336853,
|
| 1392 |
+
"learning_rate": 1.5646948111000756e-05,
|
| 1393 |
+
"loss": 0.1344,
|
| 1394 |
+
"step": 1680
|
| 1395 |
+
},
|
| 1396 |
+
{
|
| 1397 |
+
"epoch": 0.2067784167380399,
|
| 1398 |
+
"grad_norm": 0.1990954577922821,
|
| 1399 |
+
"learning_rate": 1.5579266711712185e-05,
|
| 1400 |
+
"loss": 0.1384,
|
| 1401 |
+
"step": 1690
|
| 1402 |
+
},
|
| 1403 |
+
{
|
| 1404 |
+
"epoch": 0.2080019576654839,
|
| 1405 |
+
"grad_norm": 0.14277924597263336,
|
| 1406 |
+
"learning_rate": 1.551121217159397e-05,
|
| 1407 |
+
"loss": 0.1251,
|
| 1408 |
+
"step": 1700
|
| 1409 |
+
},
|
| 1410 |
+
{
|
| 1411 |
+
"epoch": 0.2080019576654839,
|
| 1412 |
+
"eval_accuracy": 0.97199,
|
| 1413 |
+
"eval_f1_weighted": 0.9719965673787706,
|
| 1414 |
+
"eval_loss": 0.025022055953741074,
|
| 1415 |
+
"eval_macro_f1": 0.9719444578845672,
|
| 1416 |
+
"eval_precision": 0.9720688283360513,
|
| 1417 |
+
"eval_recall": 0.97199,
|
| 1418 |
+
"eval_runtime": 661.2104,
|
| 1419 |
+
"eval_samples_per_second": 151.238,
|
| 1420 |
+
"eval_steps_per_second": 1.183,
|
| 1421 |
+
"step": 1700
|
| 1422 |
+
},
|
| 1423 |
+
{
|
| 1424 |
+
"epoch": 0.20922549859292794,
|
| 1425 |
+
"grad_norm": 0.13305141031742096,
|
| 1426 |
+
"learning_rate": 1.5442789042127306e-05,
|
| 1427 |
+
"loss": 0.1465,
|
| 1428 |
+
"step": 1710
|
| 1429 |
+
},
|
| 1430 |
+
{
|
| 1431 |
+
"epoch": 0.21044903952037194,
|
| 1432 |
+
"grad_norm": 0.2761312425136566,
|
| 1433 |
+
"learning_rate": 1.5374001899444624e-05,
|
| 1434 |
+
"loss": 0.1413,
|
| 1435 |
+
"step": 1720
|
| 1436 |
+
},
|
| 1437 |
+
{
|
| 1438 |
+
"epoch": 0.21167258044781598,
|
| 1439 |
+
"grad_norm": 0.11722280085086823,
|
| 1440 |
+
"learning_rate": 1.530485534402351e-05,
|
| 1441 |
+
"loss": 0.1337,
|
| 1442 |
+
"step": 1730
|
| 1443 |
+
},
|
| 1444 |
+
{
|
| 1445 |
+
"epoch": 0.21289612137526,
|
| 1446 |
+
"grad_norm": 0.0893188863992691,
|
| 1447 |
+
"learning_rate": 1.523535400037905e-05,
|
| 1448 |
+
"loss": 0.1458,
|
| 1449 |
+
"step": 1740
|
| 1450 |
+
},
|
| 1451 |
+
{
|
| 1452 |
+
"epoch": 0.21411966230270402,
|
| 1453 |
+
"grad_norm": 0.14586731791496277,
|
| 1454 |
+
"learning_rate": 1.5165502516754526e-05,
|
| 1455 |
+
"loss": 0.1389,
|
| 1456 |
+
"step": 1750
|
| 1457 |
+
},
|
| 1458 |
+
{
|
| 1459 |
+
"epoch": 0.21534320323014805,
|
| 1460 |
+
"grad_norm": 0.23379822075366974,
|
| 1461 |
+
"learning_rate": 1.509530556481056e-05,
|
| 1462 |
+
"loss": 0.1467,
|
| 1463 |
+
"step": 1760
|
| 1464 |
+
},
|
| 1465 |
+
{
|
| 1466 |
+
"epoch": 0.21656674415759208,
|
| 1467 |
+
"grad_norm": 0.12035554647445679,
|
| 1468 |
+
"learning_rate": 1.5024767839312642e-05,
|
| 1469 |
+
"loss": 0.1386,
|
| 1470 |
+
"step": 1770
|
| 1471 |
+
},
|
| 1472 |
+
{
|
| 1473 |
+
"epoch": 0.2177902850850361,
|
| 1474 |
+
"grad_norm": 0.1451006382703781,
|
| 1475 |
+
"learning_rate": 1.495389405781719e-05,
|
| 1476 |
+
"loss": 0.1435,
|
| 1477 |
+
"step": 1780
|
| 1478 |
+
},
|
| 1479 |
+
{
|
| 1480 |
+
"epoch": 0.21901382601248012,
|
| 1481 |
+
"grad_norm": 0.08014333248138428,
|
| 1482 |
+
"learning_rate": 1.4882688960355994e-05,
|
| 1483 |
+
"loss": 0.1491,
|
| 1484 |
+
"step": 1790
|
| 1485 |
+
},
|
| 1486 |
+
{
|
| 1487 |
+
"epoch": 0.22023736693992413,
|
| 1488 |
+
"grad_norm": 0.17279869318008423,
|
| 1489 |
+
"learning_rate": 1.4811157309119235e-05,
|
| 1490 |
+
"loss": 0.151,
|
| 1491 |
+
"step": 1800
|
| 1492 |
+
},
|
| 1493 |
+
{
|
| 1494 |
+
"epoch": 0.22023736693992413,
|
| 1495 |
+
"eval_accuracy": 0.96674,
|
| 1496 |
+
"eval_f1_weighted": 0.9667576374640046,
|
| 1497 |
+
"eval_loss": 0.02695448324084282,
|
| 1498 |
+
"eval_macro_f1": 0.9667205101995933,
|
| 1499 |
+
"eval_precision": 0.9677025757044254,
|
| 1500 |
+
"eval_recall": 0.96674,
|
| 1501 |
+
"eval_runtime": 661.5931,
|
| 1502 |
+
"eval_samples_per_second": 151.15,
|
| 1503 |
+
"eval_steps_per_second": 1.182,
|
| 1504 |
+
"step": 1800
|
| 1505 |
+
},
|
| 1506 |
+
{
|
| 1507 |
+
"epoch": 0.22146090786736816,
|
| 1508 |
+
"grad_norm": 0.32028236985206604,
|
| 1509 |
+
"learning_rate": 1.473930388813698e-05,
|
| 1510 |
+
"loss": 0.1342,
|
| 1511 |
+
"step": 1810
|
| 1512 |
+
},
|
| 1513 |
+
{
|
| 1514 |
+
"epoch": 0.2226844487948122,
|
| 1515 |
+
"grad_norm": 0.2838447690010071,
|
| 1516 |
+
"learning_rate": 1.466713350295922e-05,
|
| 1517 |
+
"loss": 0.1346,
|
| 1518 |
+
"step": 1820
|
| 1519 |
+
},
|
| 1520 |
+
{
|
| 1521 |
+
"epoch": 0.2239079897222562,
|
| 1522 |
+
"grad_norm": 0.2165336161851883,
|
| 1523 |
+
"learning_rate": 1.4594650980334492e-05,
|
| 1524 |
+
"loss": 0.1447,
|
| 1525 |
+
"step": 1830
|
| 1526 |
+
},
|
| 1527 |
+
{
|
| 1528 |
+
"epoch": 0.22513153064970023,
|
| 1529 |
+
"grad_norm": 0.1966007500886917,
|
| 1530 |
+
"learning_rate": 1.4521861167887043e-05,
|
| 1531 |
+
"loss": 0.1405,
|
| 1532 |
+
"step": 1840
|
| 1533 |
+
},
|
| 1534 |
+
{
|
| 1535 |
+
"epoch": 0.22635507157714427,
|
| 1536 |
+
"grad_norm": 0.16725124418735504,
|
| 1537 |
+
"learning_rate": 1.4448768933792644e-05,
|
| 1538 |
+
"loss": 0.1272,
|
| 1539 |
+
"step": 1850
|
| 1540 |
+
},
|
| 1541 |
+
{
|
| 1542 |
+
"epoch": 0.22757861250458827,
|
| 1543 |
+
"grad_norm": 0.22608086466789246,
|
| 1544 |
+
"learning_rate": 1.4375379166452991e-05,
|
| 1545 |
+
"loss": 0.1305,
|
| 1546 |
+
"step": 1860
|
| 1547 |
+
},
|
| 1548 |
+
{
|
| 1549 |
+
"epoch": 0.2288021534320323,
|
| 1550 |
+
"grad_norm": 0.24929635226726532,
|
| 1551 |
+
"learning_rate": 1.4301696774168782e-05,
|
| 1552 |
+
"loss": 0.1276,
|
| 1553 |
+
"step": 1870
|
| 1554 |
+
},
|
| 1555 |
+
{
|
| 1556 |
+
"epoch": 0.2300256943594763,
|
| 1557 |
+
"grad_norm": 0.0894479975104332,
|
| 1558 |
+
"learning_rate": 1.4227726684811439e-05,
|
| 1559 |
+
"loss": 0.1515,
|
| 1560 |
+
"step": 1880
|
| 1561 |
+
},
|
| 1562 |
+
{
|
| 1563 |
+
"epoch": 0.23124923528692035,
|
| 1564 |
+
"grad_norm": 0.31293338537216187,
|
| 1565 |
+
"learning_rate": 1.4153473845493531e-05,
|
| 1566 |
+
"loss": 0.1262,
|
| 1567 |
+
"step": 1890
|
| 1568 |
+
},
|
| 1569 |
+
{
|
| 1570 |
+
"epoch": 0.23247277621436438,
|
| 1571 |
+
"grad_norm": 0.34327465295791626,
|
| 1572 |
+
"learning_rate": 1.4078943222237933e-05,
|
| 1573 |
+
"loss": 0.1371,
|
| 1574 |
+
"step": 1900
|
| 1575 |
+
},
|
| 1576 |
+
{
|
| 1577 |
+
"epoch": 0.23247277621436438,
|
| 1578 |
+
"eval_accuracy": 0.97149,
|
| 1579 |
+
"eval_f1_weighted": 0.9715031773882226,
|
| 1580 |
+
"eval_loss": 0.02571570873260498,
|
| 1581 |
+
"eval_macro_f1": 0.9714618804747581,
|
| 1582 |
+
"eval_precision": 0.9719368831197501,
|
| 1583 |
+
"eval_recall": 0.97149,
|
| 1584 |
+
"eval_runtime": 660.9647,
|
| 1585 |
+
"eval_samples_per_second": 151.294,
|
| 1586 |
+
"eval_steps_per_second": 1.183,
|
| 1587 |
+
"step": 1900
|
| 1588 |
+
},
|
| 1589 |
+
{
|
| 1590 |
+
"epoch": 0.23369631714180839,
|
| 1591 |
+
"grad_norm": 0.07504434138536453,
|
| 1592 |
+
"learning_rate": 1.4004139799645669e-05,
|
| 1593 |
+
"loss": 0.1506,
|
| 1594 |
+
"step": 1910
|
| 1595 |
+
},
|
| 1596 |
+
{
|
| 1597 |
+
"epoch": 0.23491985806925242,
|
| 1598 |
+
"grad_norm": 0.2954120934009552,
|
| 1599 |
+
"learning_rate": 1.3929068580562561e-05,
|
| 1600 |
+
"loss": 0.1329,
|
| 1601 |
+
"step": 1920
|
| 1602 |
+
},
|
| 1603 |
+
{
|
| 1604 |
+
"epoch": 0.23614339899669645,
|
| 1605 |
+
"grad_norm": 0.17359323799610138,
|
| 1606 |
+
"learning_rate": 1.3853734585744642e-05,
|
| 1607 |
+
"loss": 0.1347,
|
| 1608 |
+
"step": 1930
|
| 1609 |
+
},
|
| 1610 |
+
{
|
| 1611 |
+
"epoch": 0.23736693992414046,
|
| 1612 |
+
"grad_norm": 0.15386463701725006,
|
| 1613 |
+
"learning_rate": 1.3778142853522352e-05,
|
| 1614 |
+
"loss": 0.1429,
|
| 1615 |
+
"step": 1940
|
| 1616 |
+
},
|
| 1617 |
+
{
|
| 1618 |
+
"epoch": 0.2385904808515845,
|
| 1619 |
+
"grad_norm": 0.2623615264892578,
|
| 1620 |
+
"learning_rate": 1.3702298439463589e-05,
|
| 1621 |
+
"loss": 0.1296,
|
| 1622 |
+
"step": 1950
|
| 1623 |
+
},
|
| 1624 |
+
{
|
| 1625 |
+
"epoch": 0.2398140217790285,
|
| 1626 |
+
"grad_norm": 0.12422071397304535,
|
| 1627 |
+
"learning_rate": 1.3626206416035589e-05,
|
| 1628 |
+
"loss": 0.1441,
|
| 1629 |
+
"step": 1960
|
| 1630 |
+
},
|
| 1631 |
+
{
|
| 1632 |
+
"epoch": 0.24103756270647253,
|
| 1633 |
+
"grad_norm": 0.14957550168037415,
|
| 1634 |
+
"learning_rate": 1.3549871872265672e-05,
|
| 1635 |
+
"loss": 0.1406,
|
| 1636 |
+
"step": 1970
|
| 1637 |
+
},
|
| 1638 |
+
{
|
| 1639 |
+
"epoch": 0.24226110363391656,
|
| 1640 |
+
"grad_norm": 0.11441356688737869,
|
| 1641 |
+
"learning_rate": 1.3473299913400905e-05,
|
| 1642 |
+
"loss": 0.1294,
|
| 1643 |
+
"step": 1980
|
| 1644 |
+
},
|
| 1645 |
+
{
|
| 1646 |
+
"epoch": 0.24348464456136057,
|
| 1647 |
+
"grad_norm": 0.19187535345554352,
|
| 1648 |
+
"learning_rate": 1.3396495660566651e-05,
|
| 1649 |
+
"loss": 0.1427,
|
| 1650 |
+
"step": 1990
|
| 1651 |
+
},
|
| 1652 |
+
{
|
| 1653 |
+
"epoch": 0.2447081854888046,
|
| 1654 |
+
"grad_norm": 0.10940376669168472,
|
| 1655 |
+
"learning_rate": 1.3319464250424068e-05,
|
| 1656 |
+
"loss": 0.1439,
|
| 1657 |
+
"step": 2000
|
| 1658 |
+
},
|
| 1659 |
+
{
|
| 1660 |
+
"epoch": 0.2447081854888046,
|
| 1661 |
+
"eval_accuracy": 0.97729,
|
| 1662 |
+
"eval_f1_weighted": 0.9772907740157332,
|
| 1663 |
+
"eval_loss": 0.01969166286289692,
|
| 1664 |
+
"eval_macro_f1": 0.9772431978486675,
|
| 1665 |
+
"eval_precision": 0.9772926495694174,
|
| 1666 |
+
"eval_recall": 0.97729,
|
| 1667 |
+
"eval_runtime": 660.9031,
|
| 1668 |
+
"eval_samples_per_second": 151.308,
|
| 1669 |
+
"eval_steps_per_second": 1.183,
|
| 1670 |
+
"step": 2000
|
| 1671 |
+
},
|
| 1672 |
+
{
|
| 1673 |
+
"epoch": 0.24593172641624864,
|
| 1674 |
+
"grad_norm": 0.07293081283569336,
|
| 1675 |
+
"learning_rate": 1.3242210834826575e-05,
|
| 1676 |
+
"loss": 0.1322,
|
| 1677 |
+
"step": 2010
|
| 1678 |
+
},
|
| 1679 |
+
{
|
| 1680 |
+
"epoch": 0.24715526734369264,
|
| 1681 |
+
"grad_norm": 0.1582045555114746,
|
| 1682 |
+
"learning_rate": 1.31647405804753e-05,
|
| 1683 |
+
"loss": 0.1273,
|
| 1684 |
+
"step": 2020
|
| 1685 |
+
},
|
| 1686 |
+
{
|
| 1687 |
+
"epoch": 0.24837880827113668,
|
| 1688 |
+
"grad_norm": 0.29221460223197937,
|
| 1689 |
+
"learning_rate": 1.3087058668573521e-05,
|
| 1690 |
+
"loss": 0.1303,
|
| 1691 |
+
"step": 2030
|
| 1692 |
+
},
|
| 1693 |
+
{
|
| 1694 |
+
"epoch": 0.24960234919858068,
|
| 1695 |
+
"grad_norm": 0.17140211164951324,
|
| 1696 |
+
"learning_rate": 1.3009170294480149e-05,
|
| 1697 |
+
"loss": 0.1394,
|
| 1698 |
+
"step": 2040
|
| 1699 |
+
},
|
| 1700 |
+
{
|
| 1701 |
+
"epoch": 0.25082589012602474,
|
| 1702 |
+
"grad_norm": 0.3539097011089325,
|
| 1703 |
+
"learning_rate": 1.2931080667362276e-05,
|
| 1704 |
+
"loss": 0.1331,
|
| 1705 |
+
"step": 2050
|
| 1706 |
+
},
|
| 1707 |
+
{
|
| 1708 |
+
"epoch": 0.2520494310534687,
|
| 1709 |
+
"grad_norm": 0.23371531069278717,
|
| 1710 |
+
"learning_rate": 1.2852795009846775e-05,
|
| 1711 |
+
"loss": 0.1238,
|
| 1712 |
+
"step": 2060
|
| 1713 |
+
},
|
| 1714 |
+
{
|
| 1715 |
+
"epoch": 0.25327297198091275,
|
| 1716 |
+
"grad_norm": 0.21888205409049988,
|
| 1717 |
+
"learning_rate": 1.2774318557671009e-05,
|
| 1718 |
+
"loss": 0.1509,
|
| 1719 |
+
"step": 2070
|
| 1720 |
+
},
|
| 1721 |
+
{
|
| 1722 |
+
"epoch": 0.2544965129083568,
|
| 1723 |
+
"grad_norm": 0.4899176359176636,
|
| 1724 |
+
"learning_rate": 1.2695656559332671e-05,
|
| 1725 |
+
"loss": 0.1461,
|
| 1726 |
+
"step": 2080
|
| 1727 |
+
},
|
| 1728 |
+
{
|
| 1729 |
+
"epoch": 0.2557200538358008,
|
| 1730 |
+
"grad_norm": 0.1515502780675888,
|
| 1731 |
+
"learning_rate": 1.261681427573877e-05,
|
| 1732 |
+
"loss": 0.1331,
|
| 1733 |
+
"step": 2090
|
| 1734 |
+
},
|
| 1735 |
+
{
|
| 1736 |
+
"epoch": 0.25694359476324485,
|
| 1737 |
+
"grad_norm": 0.13595114648342133,
|
| 1738 |
+
"learning_rate": 1.2537796979853764e-05,
|
| 1739 |
+
"loss": 0.1484,
|
| 1740 |
+
"step": 2100
|
| 1741 |
+
},
|
| 1742 |
+
{
|
| 1743 |
+
"epoch": 0.25694359476324485,
|
| 1744 |
+
"eval_accuracy": 0.97769,
|
| 1745 |
+
"eval_f1_weighted": 0.9776954901619872,
|
| 1746 |
+
"eval_loss": 0.019225435331463814,
|
| 1747 |
+
"eval_macro_f1": 0.9776543362648145,
|
| 1748 |
+
"eval_precision": 0.9777750172865488,
|
| 1749 |
+
"eval_recall": 0.97769,
|
| 1750 |
+
"eval_runtime": 660.729,
|
| 1751 |
+
"eval_samples_per_second": 151.348,
|
| 1752 |
+
"eval_steps_per_second": 1.184,
|
| 1753 |
+
"step": 2100
|
| 1754 |
+
}
|
| 1755 |
+
],
|
| 1756 |
+
"logging_steps": 10,
|
| 1757 |
+
"max_steps": 8173,
|
| 1758 |
+
"num_input_tokens_seen": 0,
|
| 1759 |
+
"num_train_epochs": 1,
|
| 1760 |
+
"save_steps": 100,
|
| 1761 |
+
"stateful_callbacks": {
|
| 1762 |
+
"EarlyStoppingCallback": {
|
| 1763 |
+
"args": {
|
| 1764 |
+
"early_stopping_patience": 3,
|
| 1765 |
+
"early_stopping_threshold": 0.0
|
| 1766 |
+
},
|
| 1767 |
+
"attributes": {
|
| 1768 |
+
"early_stopping_patience_counter": 0
|
| 1769 |
+
}
|
| 1770 |
+
},
|
| 1771 |
+
"TrainerControl": {
|
| 1772 |
+
"args": {
|
| 1773 |
+
"should_epoch_stop": false,
|
| 1774 |
+
"should_evaluate": false,
|
| 1775 |
+
"should_log": false,
|
| 1776 |
+
"should_save": true,
|
| 1777 |
+
"should_training_stop": false
|
| 1778 |
+
},
|
| 1779 |
+
"attributes": {}
|
| 1780 |
+
}
|
| 1781 |
+
},
|
| 1782 |
+
"total_flos": 3.564604425568589e+16,
|
| 1783 |
+
"train_batch_size": 64,
|
| 1784 |
+
"trial_name": null,
|
| 1785 |
+
"trial_params": null
|
| 1786 |
+
}
|
checkpoint-2100/training_args.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:34af5433a3002793dc6233ee444a7f3918508304d82b9638eca4b9f53208c9bf
|
| 3 |
+
size 5841
|
checkpoint-2100/vocab.json
ADDED
|
The diff for this file is too large to render.
See raw diff
|
|
|
checkpoint-2200/config.json
ADDED
|
@@ -0,0 +1,29 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"architectures": [
|
| 3 |
+
"RobertaForSequenceClassification"
|
| 4 |
+
],
|
| 5 |
+
"attention_probs_dropout_prob": 0.1,
|
| 6 |
+
"bos_token_id": 0,
|
| 7 |
+
"classifier_dropout": null,
|
| 8 |
+
"dtype": "float32",
|
| 9 |
+
"eos_token_id": 2,
|
| 10 |
+
"gradient_checkpointing": false,
|
| 11 |
+
"hidden_act": "gelu",
|
| 12 |
+
"hidden_dropout_prob": 0.1,
|
| 13 |
+
"hidden_size": 768,
|
| 14 |
+
"initializer_range": 0.02,
|
| 15 |
+
"intermediate_size": 3072,
|
| 16 |
+
"layer_norm_eps": 1e-05,
|
| 17 |
+
"max_position_embeddings": 1026,
|
| 18 |
+
"model_type": "roberta",
|
| 19 |
+
"num_attention_heads": 12,
|
| 20 |
+
"num_hidden_layers": 12,
|
| 21 |
+
"output_past": true,
|
| 22 |
+
"pad_token_id": 1,
|
| 23 |
+
"position_embedding_type": "absolute",
|
| 24 |
+
"problem_type": "single_label_classification",
|
| 25 |
+
"transformers_version": "4.56.0",
|
| 26 |
+
"type_vocab_size": 10,
|
| 27 |
+
"use_cache": true,
|
| 28 |
+
"vocab_size": 51416
|
| 29 |
+
}
|
checkpoint-2200/merges.txt
ADDED
|
The diff for this file is too large to render.
See raw diff
|
|
|
checkpoint-2200/model.safetensors
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:335df2fb9c96b5e579051e5a801512ed4e3d23ef053139b2fb42f7445f061aaa
|
| 3 |
+
size 506505672
|
checkpoint-2200/optimizer.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:5bcdee09dd93cdd197e6ea1eee09b9979eaf49d3baada43d87bee36a080a9c8d
|
| 3 |
+
size 1007622411
|
checkpoint-2200/rng_state.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:15c0d4180bda455330407ed0bb58739c7ec87b45b8bb6d957f4d11a53fa4cc59
|
| 3 |
+
size 14645
|
checkpoint-2200/scaler.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:90bbc0bb51ada144c607a2d0e8dae7a41e3cca0fcbc5a9257404d9d17b7a3f9d
|
| 3 |
+
size 1383
|
checkpoint-2200/scheduler.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:a7f7f10addafb465c38e374d351212daa821676c6da17d02f7fb0ffb64d30980
|
| 3 |
+
size 1977
|
checkpoint-2200/special_tokens_map.json
ADDED
|
@@ -0,0 +1,51 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"bos_token": {
|
| 3 |
+
"content": "<s>",
|
| 4 |
+
"lstrip": false,
|
| 5 |
+
"normalized": true,
|
| 6 |
+
"rstrip": false,
|
| 7 |
+
"single_word": false
|
| 8 |
+
},
|
| 9 |
+
"cls_token": {
|
| 10 |
+
"content": "<s>",
|
| 11 |
+
"lstrip": false,
|
| 12 |
+
"normalized": true,
|
| 13 |
+
"rstrip": false,
|
| 14 |
+
"single_word": false
|
| 15 |
+
},
|
| 16 |
+
"eos_token": {
|
| 17 |
+
"content": "</s>",
|
| 18 |
+
"lstrip": false,
|
| 19 |
+
"normalized": true,
|
| 20 |
+
"rstrip": false,
|
| 21 |
+
"single_word": false
|
| 22 |
+
},
|
| 23 |
+
"mask_token": {
|
| 24 |
+
"content": "<mask>",
|
| 25 |
+
"lstrip": true,
|
| 26 |
+
"normalized": true,
|
| 27 |
+
"rstrip": false,
|
| 28 |
+
"single_word": false
|
| 29 |
+
},
|
| 30 |
+
"pad_token": {
|
| 31 |
+
"content": "<pad>",
|
| 32 |
+
"lstrip": false,
|
| 33 |
+
"normalized": true,
|
| 34 |
+
"rstrip": false,
|
| 35 |
+
"single_word": false
|
| 36 |
+
},
|
| 37 |
+
"sep_token": {
|
| 38 |
+
"content": "</s>",
|
| 39 |
+
"lstrip": false,
|
| 40 |
+
"normalized": true,
|
| 41 |
+
"rstrip": false,
|
| 42 |
+
"single_word": false
|
| 43 |
+
},
|
| 44 |
+
"unk_token": {
|
| 45 |
+
"content": "<unk>",
|
| 46 |
+
"lstrip": false,
|
| 47 |
+
"normalized": true,
|
| 48 |
+
"rstrip": false,
|
| 49 |
+
"single_word": false
|
| 50 |
+
}
|
| 51 |
+
}
|
checkpoint-2200/tokenizer.json
ADDED
|
The diff for this file is too large to render.
See raw diff
|
|
|
checkpoint-2200/tokenizer_config.json
ADDED
|
@@ -0,0 +1,58 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"add_prefix_space": false,
|
| 3 |
+
"added_tokens_decoder": {
|
| 4 |
+
"0": {
|
| 5 |
+
"content": "<s>",
|
| 6 |
+
"lstrip": false,
|
| 7 |
+
"normalized": true,
|
| 8 |
+
"rstrip": false,
|
| 9 |
+
"single_word": false,
|
| 10 |
+
"special": true
|
| 11 |
+
},
|
| 12 |
+
"1": {
|
| 13 |
+
"content": "<pad>",
|
| 14 |
+
"lstrip": false,
|
| 15 |
+
"normalized": true,
|
| 16 |
+
"rstrip": false,
|
| 17 |
+
"single_word": false,
|
| 18 |
+
"special": true
|
| 19 |
+
},
|
| 20 |
+
"2": {
|
| 21 |
+
"content": "</s>",
|
| 22 |
+
"lstrip": false,
|
| 23 |
+
"normalized": true,
|
| 24 |
+
"rstrip": false,
|
| 25 |
+
"single_word": false,
|
| 26 |
+
"special": true
|
| 27 |
+
},
|
| 28 |
+
"3": {
|
| 29 |
+
"content": "<unk>",
|
| 30 |
+
"lstrip": false,
|
| 31 |
+
"normalized": true,
|
| 32 |
+
"rstrip": false,
|
| 33 |
+
"single_word": false,
|
| 34 |
+
"special": true
|
| 35 |
+
},
|
| 36 |
+
"4": {
|
| 37 |
+
"content": "<mask>",
|
| 38 |
+
"lstrip": true,
|
| 39 |
+
"normalized": true,
|
| 40 |
+
"rstrip": false,
|
| 41 |
+
"single_word": false,
|
| 42 |
+
"special": true
|
| 43 |
+
}
|
| 44 |
+
},
|
| 45 |
+
"bos_token": "<s>",
|
| 46 |
+
"clean_up_tokenization_spaces": false,
|
| 47 |
+
"cls_token": "<s>",
|
| 48 |
+
"eos_token": "</s>",
|
| 49 |
+
"errors": "replace",
|
| 50 |
+
"extra_special_tokens": {},
|
| 51 |
+
"mask_token": "<mask>",
|
| 52 |
+
"model_max_length": 1000000000000000019884624838656,
|
| 53 |
+
"pad_token": "<pad>",
|
| 54 |
+
"sep_token": "</s>",
|
| 55 |
+
"tokenizer_class": "RobertaTokenizer",
|
| 56 |
+
"trim_offsets": true,
|
| 57 |
+
"unk_token": "<unk>"
|
| 58 |
+
}
|
checkpoint-2200/trainer_state.json
ADDED
|
@@ -0,0 +1,1869 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_global_step": 2100,
|
| 3 |
+
"best_metric": 0.9776543362648145,
|
| 4 |
+
"best_model_checkpoint": "taskA-unixcoder-focal/checkpoint-2100",
|
| 5 |
+
"epoch": 0.2691790040376851,
|
| 6 |
+
"eval_steps": 100,
|
| 7 |
+
"global_step": 2200,
|
| 8 |
+
"is_hyper_param_search": false,
|
| 9 |
+
"is_local_process_zero": true,
|
| 10 |
+
"is_world_process_zero": true,
|
| 11 |
+
"log_history": [
|
| 12 |
+
{
|
| 13 |
+
"epoch": 0.001223540927444023,
|
| 14 |
+
"grad_norm": 0.1898110806941986,
|
| 15 |
+
"learning_rate": 3.6734693877551025e-07,
|
| 16 |
+
"loss": 0.4422,
|
| 17 |
+
"step": 10
|
| 18 |
+
},
|
| 19 |
+
{
|
| 20 |
+
"epoch": 0.002447081854888046,
|
| 21 |
+
"grad_norm": 0.3350900411605835,
|
| 22 |
+
"learning_rate": 7.346938775510205e-07,
|
| 23 |
+
"loss": 0.4323,
|
| 24 |
+
"step": 20
|
| 25 |
+
},
|
| 26 |
+
{
|
| 27 |
+
"epoch": 0.003670622782332069,
|
| 28 |
+
"grad_norm": 0.3723449110984802,
|
| 29 |
+
"learning_rate": 1.142857142857143e-06,
|
| 30 |
+
"loss": 0.4362,
|
| 31 |
+
"step": 30
|
| 32 |
+
},
|
| 33 |
+
{
|
| 34 |
+
"epoch": 0.004894163709776092,
|
| 35 |
+
"grad_norm": 0.3299769461154938,
|
| 36 |
+
"learning_rate": 1.5510204081632654e-06,
|
| 37 |
+
"loss": 0.4191,
|
| 38 |
+
"step": 40
|
| 39 |
+
},
|
| 40 |
+
{
|
| 41 |
+
"epoch": 0.006117704637220115,
|
| 42 |
+
"grad_norm": 0.26681044697761536,
|
| 43 |
+
"learning_rate": 1.9591836734693877e-06,
|
| 44 |
+
"loss": 0.4289,
|
| 45 |
+
"step": 50
|
| 46 |
+
},
|
| 47 |
+
{
|
| 48 |
+
"epoch": 0.007341245564664138,
|
| 49 |
+
"grad_norm": 0.30233338475227356,
|
| 50 |
+
"learning_rate": 2.3673469387755103e-06,
|
| 51 |
+
"loss": 0.4083,
|
| 52 |
+
"step": 60
|
| 53 |
+
},
|
| 54 |
+
{
|
| 55 |
+
"epoch": 0.008564786492108161,
|
| 56 |
+
"grad_norm": 0.31031334400177,
|
| 57 |
+
"learning_rate": 2.775510204081633e-06,
|
| 58 |
+
"loss": 0.404,
|
| 59 |
+
"step": 70
|
| 60 |
+
},
|
| 61 |
+
{
|
| 62 |
+
"epoch": 0.009788327419552184,
|
| 63 |
+
"grad_norm": 0.2552475035190582,
|
| 64 |
+
"learning_rate": 3.183673469387755e-06,
|
| 65 |
+
"loss": 0.3889,
|
| 66 |
+
"step": 80
|
| 67 |
+
},
|
| 68 |
+
{
|
| 69 |
+
"epoch": 0.011011868346996207,
|
| 70 |
+
"grad_norm": 0.3131052553653717,
|
| 71 |
+
"learning_rate": 3.5918367346938777e-06,
|
| 72 |
+
"loss": 0.3813,
|
| 73 |
+
"step": 90
|
| 74 |
+
},
|
| 75 |
+
{
|
| 76 |
+
"epoch": 0.01223540927444023,
|
| 77 |
+
"grad_norm": 0.3377229571342468,
|
| 78 |
+
"learning_rate": 4.000000000000001e-06,
|
| 79 |
+
"loss": 0.3727,
|
| 80 |
+
"step": 100
|
| 81 |
+
},
|
| 82 |
+
{
|
| 83 |
+
"epoch": 0.01223540927444023,
|
| 84 |
+
"eval_accuracy": 0.6961,
|
| 85 |
+
"eval_f1_weighted": 0.6955947095331424,
|
| 86 |
+
"eval_loss": 0.15026023983955383,
|
| 87 |
+
"eval_macro_f1": 0.6959280508399847,
|
| 88 |
+
"eval_precision": 0.7019396279846728,
|
| 89 |
+
"eval_recall": 0.6961,
|
| 90 |
+
"eval_runtime": 789.7648,
|
| 91 |
+
"eval_samples_per_second": 126.62,
|
| 92 |
+
"eval_steps_per_second": 0.99,
|
| 93 |
+
"step": 100
|
| 94 |
+
},
|
| 95 |
+
{
|
| 96 |
+
"epoch": 0.013458950201884253,
|
| 97 |
+
"grad_norm": 0.3366774618625641,
|
| 98 |
+
"learning_rate": 4.408163265306123e-06,
|
| 99 |
+
"loss": 0.3671,
|
| 100 |
+
"step": 110
|
| 101 |
+
},
|
| 102 |
+
{
|
| 103 |
+
"epoch": 0.014682491129328276,
|
| 104 |
+
"grad_norm": 0.33491575717926025,
|
| 105 |
+
"learning_rate": 4.816326530612245e-06,
|
| 106 |
+
"loss": 0.3498,
|
| 107 |
+
"step": 120
|
| 108 |
+
},
|
| 109 |
+
{
|
| 110 |
+
"epoch": 0.0159060320567723,
|
| 111 |
+
"grad_norm": 0.4119006097316742,
|
| 112 |
+
"learning_rate": 5.2244897959183684e-06,
|
| 113 |
+
"loss": 0.3409,
|
| 114 |
+
"step": 130
|
| 115 |
+
},
|
| 116 |
+
{
|
| 117 |
+
"epoch": 0.017129572984216322,
|
| 118 |
+
"grad_norm": 0.2964725196361542,
|
| 119 |
+
"learning_rate": 5.63265306122449e-06,
|
| 120 |
+
"loss": 0.3433,
|
| 121 |
+
"step": 140
|
| 122 |
+
},
|
| 123 |
+
{
|
| 124 |
+
"epoch": 0.018353113911660345,
|
| 125 |
+
"grad_norm": 0.3075334429740906,
|
| 126 |
+
"learning_rate": 6.040816326530613e-06,
|
| 127 |
+
"loss": 0.3298,
|
| 128 |
+
"step": 150
|
| 129 |
+
},
|
| 130 |
+
{
|
| 131 |
+
"epoch": 0.019576654839104368,
|
| 132 |
+
"grad_norm": 0.33718687295913696,
|
| 133 |
+
"learning_rate": 6.4489795918367345e-06,
|
| 134 |
+
"loss": 0.3297,
|
| 135 |
+
"step": 160
|
| 136 |
+
},
|
| 137 |
+
{
|
| 138 |
+
"epoch": 0.02080019576654839,
|
| 139 |
+
"grad_norm": 0.2809618413448334,
|
| 140 |
+
"learning_rate": 6.857142857142858e-06,
|
| 141 |
+
"loss": 0.3275,
|
| 142 |
+
"step": 170
|
| 143 |
+
},
|
| 144 |
+
{
|
| 145 |
+
"epoch": 0.022023736693992414,
|
| 146 |
+
"grad_norm": 0.37805014848709106,
|
| 147 |
+
"learning_rate": 7.2653061224489805e-06,
|
| 148 |
+
"loss": 0.3201,
|
| 149 |
+
"step": 180
|
| 150 |
+
},
|
| 151 |
+
{
|
| 152 |
+
"epoch": 0.023247277621436437,
|
| 153 |
+
"grad_norm": 0.2940429151058197,
|
| 154 |
+
"learning_rate": 7.673469387755102e-06,
|
| 155 |
+
"loss": 0.3042,
|
| 156 |
+
"step": 190
|
| 157 |
+
},
|
| 158 |
+
{
|
| 159 |
+
"epoch": 0.02447081854888046,
|
| 160 |
+
"grad_norm": 0.3286060690879822,
|
| 161 |
+
"learning_rate": 8.081632653061225e-06,
|
| 162 |
+
"loss": 0.3071,
|
| 163 |
+
"step": 200
|
| 164 |
+
},
|
| 165 |
+
{
|
| 166 |
+
"epoch": 0.02447081854888046,
|
| 167 |
+
"eval_accuracy": 0.77807,
|
| 168 |
+
"eval_f1_weighted": 0.7781129395311969,
|
| 169 |
+
"eval_loss": 0.12153773754835129,
|
| 170 |
+
"eval_macro_f1": 0.7780651598452646,
|
| 171 |
+
"eval_precision": 0.7808306361972198,
|
| 172 |
+
"eval_recall": 0.77807,
|
| 173 |
+
"eval_runtime": 792.7157,
|
| 174 |
+
"eval_samples_per_second": 126.149,
|
| 175 |
+
"eval_steps_per_second": 0.986,
|
| 176 |
+
"step": 200
|
| 177 |
+
},
|
| 178 |
+
{
|
| 179 |
+
"epoch": 0.025694359476324483,
|
| 180 |
+
"grad_norm": 0.33857154846191406,
|
| 181 |
+
"learning_rate": 8.489795918367347e-06,
|
| 182 |
+
"loss": 0.2964,
|
| 183 |
+
"step": 210
|
| 184 |
+
},
|
| 185 |
+
{
|
| 186 |
+
"epoch": 0.026917900403768506,
|
| 187 |
+
"grad_norm": 0.2752092182636261,
|
| 188 |
+
"learning_rate": 8.89795918367347e-06,
|
| 189 |
+
"loss": 0.3,
|
| 190 |
+
"step": 220
|
| 191 |
+
},
|
| 192 |
+
{
|
| 193 |
+
"epoch": 0.02814144133121253,
|
| 194 |
+
"grad_norm": 0.4952544569969177,
|
| 195 |
+
"learning_rate": 9.306122448979593e-06,
|
| 196 |
+
"loss": 0.294,
|
| 197 |
+
"step": 230
|
| 198 |
+
},
|
| 199 |
+
{
|
| 200 |
+
"epoch": 0.029364982258656552,
|
| 201 |
+
"grad_norm": 0.37926042079925537,
|
| 202 |
+
"learning_rate": 9.714285714285715e-06,
|
| 203 |
+
"loss": 0.2964,
|
| 204 |
+
"step": 240
|
| 205 |
+
},
|
| 206 |
+
{
|
| 207 |
+
"epoch": 0.030588523186100575,
|
| 208 |
+
"grad_norm": 0.3222455680370331,
|
| 209 |
+
"learning_rate": 1.0122448979591836e-05,
|
| 210 |
+
"loss": 0.283,
|
| 211 |
+
"step": 250
|
| 212 |
+
},
|
| 213 |
+
{
|
| 214 |
+
"epoch": 0.0318120641135446,
|
| 215 |
+
"grad_norm": 0.2930658459663391,
|
| 216 |
+
"learning_rate": 1.053061224489796e-05,
|
| 217 |
+
"loss": 0.2793,
|
| 218 |
+
"step": 260
|
| 219 |
+
},
|
| 220 |
+
{
|
| 221 |
+
"epoch": 0.03303560504098862,
|
| 222 |
+
"grad_norm": 0.4313414692878723,
|
| 223 |
+
"learning_rate": 1.0938775510204083e-05,
|
| 224 |
+
"loss": 0.2728,
|
| 225 |
+
"step": 270
|
| 226 |
+
},
|
| 227 |
+
{
|
| 228 |
+
"epoch": 0.034259145968432644,
|
| 229 |
+
"grad_norm": 0.37351569533348083,
|
| 230 |
+
"learning_rate": 1.1346938775510206e-05,
|
| 231 |
+
"loss": 0.2752,
|
| 232 |
+
"step": 280
|
| 233 |
+
},
|
| 234 |
+
{
|
| 235 |
+
"epoch": 0.035482686895876664,
|
| 236 |
+
"grad_norm": 0.22772285342216492,
|
| 237 |
+
"learning_rate": 1.1755102040816328e-05,
|
| 238 |
+
"loss": 0.274,
|
| 239 |
+
"step": 290
|
| 240 |
+
},
|
| 241 |
+
{
|
| 242 |
+
"epoch": 0.03670622782332069,
|
| 243 |
+
"grad_norm": 0.15888753533363342,
|
| 244 |
+
"learning_rate": 1.2163265306122449e-05,
|
| 245 |
+
"loss": 0.264,
|
| 246 |
+
"step": 300
|
| 247 |
+
},
|
| 248 |
+
{
|
| 249 |
+
"epoch": 0.03670622782332069,
|
| 250 |
+
"eval_accuracy": 0.82326,
|
| 251 |
+
"eval_f1_weighted": 0.8227034206790359,
|
| 252 |
+
"eval_loss": 0.10234300792217255,
|
| 253 |
+
"eval_macro_f1": 0.8230100448030374,
|
| 254 |
+
"eval_precision": 0.8338333810964687,
|
| 255 |
+
"eval_recall": 0.82326,
|
| 256 |
+
"eval_runtime": 789.0168,
|
| 257 |
+
"eval_samples_per_second": 126.74,
|
| 258 |
+
"eval_steps_per_second": 0.991,
|
| 259 |
+
"step": 300
|
| 260 |
+
},
|
| 261 |
+
{
|
| 262 |
+
"epoch": 0.03792976875076471,
|
| 263 |
+
"grad_norm": 0.23560188710689545,
|
| 264 |
+
"learning_rate": 1.2571428571428572e-05,
|
| 265 |
+
"loss": 0.2753,
|
| 266 |
+
"step": 310
|
| 267 |
+
},
|
| 268 |
+
{
|
| 269 |
+
"epoch": 0.039153309678208736,
|
| 270 |
+
"grad_norm": 0.18699614703655243,
|
| 271 |
+
"learning_rate": 1.2979591836734696e-05,
|
| 272 |
+
"loss": 0.2565,
|
| 273 |
+
"step": 320
|
| 274 |
+
},
|
| 275 |
+
{
|
| 276 |
+
"epoch": 0.040376850605652756,
|
| 277 |
+
"grad_norm": 0.24035991728305817,
|
| 278 |
+
"learning_rate": 1.3387755102040818e-05,
|
| 279 |
+
"loss": 0.2716,
|
| 280 |
+
"step": 330
|
| 281 |
+
},
|
| 282 |
+
{
|
| 283 |
+
"epoch": 0.04160039153309678,
|
| 284 |
+
"grad_norm": 0.15309660136699677,
|
| 285 |
+
"learning_rate": 1.3795918367346941e-05,
|
| 286 |
+
"loss": 0.2515,
|
| 287 |
+
"step": 340
|
| 288 |
+
},
|
| 289 |
+
{
|
| 290 |
+
"epoch": 0.0428239324605408,
|
| 291 |
+
"grad_norm": 0.3382055461406708,
|
| 292 |
+
"learning_rate": 1.4204081632653062e-05,
|
| 293 |
+
"loss": 0.2597,
|
| 294 |
+
"step": 350
|
| 295 |
+
},
|
| 296 |
+
{
|
| 297 |
+
"epoch": 0.04404747338798483,
|
| 298 |
+
"grad_norm": 0.24439455568790436,
|
| 299 |
+
"learning_rate": 1.4612244897959185e-05,
|
| 300 |
+
"loss": 0.247,
|
| 301 |
+
"step": 360
|
| 302 |
+
},
|
| 303 |
+
{
|
| 304 |
+
"epoch": 0.04527101431542885,
|
| 305 |
+
"grad_norm": 0.26233720779418945,
|
| 306 |
+
"learning_rate": 1.5020408163265307e-05,
|
| 307 |
+
"loss": 0.2431,
|
| 308 |
+
"step": 370
|
| 309 |
+
},
|
| 310 |
+
{
|
| 311 |
+
"epoch": 0.046494555242872875,
|
| 312 |
+
"grad_norm": 0.20477451384067535,
|
| 313 |
+
"learning_rate": 1.5387755102040818e-05,
|
| 314 |
+
"loss": 0.2538,
|
| 315 |
+
"step": 380
|
| 316 |
+
},
|
| 317 |
+
{
|
| 318 |
+
"epoch": 0.047718096170316894,
|
| 319 |
+
"grad_norm": 0.2765439748764038,
|
| 320 |
+
"learning_rate": 1.5795918367346942e-05,
|
| 321 |
+
"loss": 0.2404,
|
| 322 |
+
"step": 390
|
| 323 |
+
},
|
| 324 |
+
{
|
| 325 |
+
"epoch": 0.04894163709776092,
|
| 326 |
+
"grad_norm": 0.13196279108524323,
|
| 327 |
+
"learning_rate": 1.6204081632653063e-05,
|
| 328 |
+
"loss": 0.2393,
|
| 329 |
+
"step": 400
|
| 330 |
+
},
|
| 331 |
+
{
|
| 332 |
+
"epoch": 0.04894163709776092,
|
| 333 |
+
"eval_accuracy": 0.84563,
|
| 334 |
+
"eval_f1_weighted": 0.8448128964509274,
|
| 335 |
+
"eval_loss": 0.09028046578168869,
|
| 336 |
+
"eval_macro_f1": 0.8451923512067675,
|
| 337 |
+
"eval_precision": 0.861024732271907,
|
| 338 |
+
"eval_recall": 0.84563,
|
| 339 |
+
"eval_runtime": 787.6067,
|
| 340 |
+
"eval_samples_per_second": 126.967,
|
| 341 |
+
"eval_steps_per_second": 0.993,
|
| 342 |
+
"step": 400
|
| 343 |
+
},
|
| 344 |
+
{
|
| 345 |
+
"epoch": 0.05016517802520494,
|
| 346 |
+
"grad_norm": 0.0924634337425232,
|
| 347 |
+
"learning_rate": 1.6612244897959184e-05,
|
| 348 |
+
"loss": 0.2331,
|
| 349 |
+
"step": 410
|
| 350 |
+
},
|
| 351 |
+
{
|
| 352 |
+
"epoch": 0.05138871895264897,
|
| 353 |
+
"grad_norm": 0.277784138917923,
|
| 354 |
+
"learning_rate": 1.7020408163265308e-05,
|
| 355 |
+
"loss": 0.2301,
|
| 356 |
+
"step": 420
|
| 357 |
+
},
|
| 358 |
+
{
|
| 359 |
+
"epoch": 0.052612259880092986,
|
| 360 |
+
"grad_norm": 0.32299959659576416,
|
| 361 |
+
"learning_rate": 1.742857142857143e-05,
|
| 362 |
+
"loss": 0.2218,
|
| 363 |
+
"step": 430
|
| 364 |
+
},
|
| 365 |
+
{
|
| 366 |
+
"epoch": 0.05383580080753701,
|
| 367 |
+
"grad_norm": 0.18170586228370667,
|
| 368 |
+
"learning_rate": 1.7836734693877553e-05,
|
| 369 |
+
"loss": 0.2318,
|
| 370 |
+
"step": 440
|
| 371 |
+
},
|
| 372 |
+
{
|
| 373 |
+
"epoch": 0.05505934173498103,
|
| 374 |
+
"grad_norm": 0.1520216166973114,
|
| 375 |
+
"learning_rate": 1.8244897959183674e-05,
|
| 376 |
+
"loss": 0.2136,
|
| 377 |
+
"step": 450
|
| 378 |
+
},
|
| 379 |
+
{
|
| 380 |
+
"epoch": 0.05628288266242506,
|
| 381 |
+
"grad_norm": 0.23452645540237427,
|
| 382 |
+
"learning_rate": 1.8653061224489795e-05,
|
| 383 |
+
"loss": 0.2219,
|
| 384 |
+
"step": 460
|
| 385 |
+
},
|
| 386 |
+
{
|
| 387 |
+
"epoch": 0.05750642358986908,
|
| 388 |
+
"grad_norm": 0.10036301612854004,
|
| 389 |
+
"learning_rate": 1.906122448979592e-05,
|
| 390 |
+
"loss": 0.2213,
|
| 391 |
+
"step": 470
|
| 392 |
+
},
|
| 393 |
+
{
|
| 394 |
+
"epoch": 0.058729964517313105,
|
| 395 |
+
"grad_norm": 0.24100220203399658,
|
| 396 |
+
"learning_rate": 1.9469387755102044e-05,
|
| 397 |
+
"loss": 0.2079,
|
| 398 |
+
"step": 480
|
| 399 |
+
},
|
| 400 |
+
{
|
| 401 |
+
"epoch": 0.059953505444757124,
|
| 402 |
+
"grad_norm": 0.28893032670021057,
|
| 403 |
+
"learning_rate": 1.9877551020408165e-05,
|
| 404 |
+
"loss": 0.1978,
|
| 405 |
+
"step": 490
|
| 406 |
+
},
|
| 407 |
+
{
|
| 408 |
+
"epoch": 0.06117704637220115,
|
| 409 |
+
"grad_norm": 0.22918537259101868,
|
| 410 |
+
"learning_rate": 1.9999836143766126e-05,
|
| 411 |
+
"loss": 0.202,
|
| 412 |
+
"step": 500
|
| 413 |
+
},
|
| 414 |
+
{
|
| 415 |
+
"epoch": 0.06117704637220115,
|
| 416 |
+
"eval_accuracy": 0.91628,
|
| 417 |
+
"eval_f1_weighted": 0.9163231793856368,
|
| 418 |
+
"eval_loss": 0.055908624082803726,
|
| 419 |
+
"eval_macro_f1": 0.9162187423616152,
|
| 420 |
+
"eval_precision": 0.9170320825159874,
|
| 421 |
+
"eval_recall": 0.91628,
|
| 422 |
+
"eval_runtime": 760.1138,
|
| 423 |
+
"eval_samples_per_second": 131.559,
|
| 424 |
+
"eval_steps_per_second": 1.029,
|
| 425 |
+
"step": 500
|
| 426 |
+
},
|
| 427 |
+
{
|
| 428 |
+
"epoch": 0.06240058729964517,
|
| 429 |
+
"grad_norm": 0.2930404841899872,
|
| 430 |
+
"learning_rate": 1.999903359554741e-05,
|
| 431 |
+
"loss": 0.1979,
|
| 432 |
+
"step": 510
|
| 433 |
+
},
|
| 434 |
+
{
|
| 435 |
+
"epoch": 0.0636241282270892,
|
| 436 |
+
"grad_norm": 0.0889664739370346,
|
| 437 |
+
"learning_rate": 1.9997562312908153e-05,
|
| 438 |
+
"loss": 0.1953,
|
| 439 |
+
"step": 520
|
| 440 |
+
},
|
| 441 |
+
{
|
| 442 |
+
"epoch": 0.06484766915453322,
|
| 443 |
+
"grad_norm": 0.1687588095664978,
|
| 444 |
+
"learning_rate": 1.9995422394247598e-05,
|
| 445 |
+
"loss": 0.1935,
|
| 446 |
+
"step": 530
|
| 447 |
+
},
|
| 448 |
+
{
|
| 449 |
+
"epoch": 0.06607121008197724,
|
| 450 |
+
"grad_norm": 0.2399725466966629,
|
| 451 |
+
"learning_rate": 1.9992613982683296e-05,
|
| 452 |
+
"loss": 0.1818,
|
| 453 |
+
"step": 540
|
| 454 |
+
},
|
| 455 |
+
{
|
| 456 |
+
"epoch": 0.06729475100942127,
|
| 457 |
+
"grad_norm": 0.2016785889863968,
|
| 458 |
+
"learning_rate": 1.9989137266041552e-05,
|
| 459 |
+
"loss": 0.1893,
|
| 460 |
+
"step": 550
|
| 461 |
+
},
|
| 462 |
+
{
|
| 463 |
+
"epoch": 0.06851829193686529,
|
| 464 |
+
"grad_norm": 0.17420713603496552,
|
| 465 |
+
"learning_rate": 1.9984992476844845e-05,
|
| 466 |
+
"loss": 0.1895,
|
| 467 |
+
"step": 560
|
| 468 |
+
},
|
| 469 |
+
{
|
| 470 |
+
"epoch": 0.06974183286430931,
|
| 471 |
+
"grad_norm": 0.2180609554052353,
|
| 472 |
+
"learning_rate": 1.998017989229629e-05,
|
| 473 |
+
"loss": 0.1836,
|
| 474 |
+
"step": 570
|
| 475 |
+
},
|
| 476 |
+
{
|
| 477 |
+
"epoch": 0.07096537379175333,
|
| 478 |
+
"grad_norm": 0.15563039481639862,
|
| 479 |
+
"learning_rate": 1.9974699834261074e-05,
|
| 480 |
+
"loss": 0.182,
|
| 481 |
+
"step": 580
|
| 482 |
+
},
|
| 483 |
+
{
|
| 484 |
+
"epoch": 0.07218891471919736,
|
| 485 |
+
"grad_norm": 0.25410088896751404,
|
| 486 |
+
"learning_rate": 1.996855266924497e-05,
|
| 487 |
+
"loss": 0.1862,
|
| 488 |
+
"step": 590
|
| 489 |
+
},
|
| 490 |
+
{
|
| 491 |
+
"epoch": 0.07341245564664138,
|
| 492 |
+
"grad_norm": 0.24206259846687317,
|
| 493 |
+
"learning_rate": 1.996173880836979e-05,
|
| 494 |
+
"loss": 0.1818,
|
| 495 |
+
"step": 600
|
| 496 |
+
},
|
| 497 |
+
{
|
| 498 |
+
"epoch": 0.07341245564664138,
|
| 499 |
+
"eval_accuracy": 0.91633,
|
| 500 |
+
"eval_f1_weighted": 0.9162437523349878,
|
| 501 |
+
"eval_loss": 0.05625538155436516,
|
| 502 |
+
"eval_macro_f1": 0.9163072974421806,
|
| 503 |
+
"eval_precision": 0.9233764457540103,
|
| 504 |
+
"eval_recall": 0.91633,
|
| 505 |
+
"eval_runtime": 759.6815,
|
| 506 |
+
"eval_samples_per_second": 131.634,
|
| 507 |
+
"eval_steps_per_second": 1.029,
|
| 508 |
+
"step": 600
|
| 509 |
+
},
|
| 510 |
+
{
|
| 511 |
+
"epoch": 0.0746359965740854,
|
| 512 |
+
"grad_norm": 0.27251410484313965,
|
| 513 |
+
"learning_rate": 1.9954258707345903e-05,
|
| 514 |
+
"loss": 0.1845,
|
| 515 |
+
"step": 610
|
| 516 |
+
},
|
| 517 |
+
{
|
| 518 |
+
"epoch": 0.07585953750152942,
|
| 519 |
+
"grad_norm": 0.1545839011669159,
|
| 520 |
+
"learning_rate": 1.9946112866441757e-05,
|
| 521 |
+
"loss": 0.1861,
|
| 522 |
+
"step": 620
|
| 523 |
+
},
|
| 524 |
+
{
|
| 525 |
+
"epoch": 0.07708307842897345,
|
| 526 |
+
"grad_norm": 0.2535873055458069,
|
| 527 |
+
"learning_rate": 1.9937301830450423e-05,
|
| 528 |
+
"loss": 0.1772,
|
| 529 |
+
"step": 630
|
| 530 |
+
},
|
| 531 |
+
{
|
| 532 |
+
"epoch": 0.07830661935641747,
|
| 533 |
+
"grad_norm": 0.2617948651313782,
|
| 534 |
+
"learning_rate": 1.992782618865315e-05,
|
| 535 |
+
"loss": 0.1711,
|
| 536 |
+
"step": 640
|
| 537 |
+
},
|
| 538 |
+
{
|
| 539 |
+
"epoch": 0.07953016028386149,
|
| 540 |
+
"grad_norm": 0.38676971197128296,
|
| 541 |
+
"learning_rate": 1.9917686574779972e-05,
|
| 542 |
+
"loss": 0.1748,
|
| 543 |
+
"step": 650
|
| 544 |
+
},
|
| 545 |
+
{
|
| 546 |
+
"epoch": 0.08075370121130551,
|
| 547 |
+
"grad_norm": 0.1511957049369812,
|
| 548 |
+
"learning_rate": 1.9906883666967296e-05,
|
| 549 |
+
"loss": 0.1825,
|
| 550 |
+
"step": 660
|
| 551 |
+
},
|
| 552 |
+
{
|
| 553 |
+
"epoch": 0.08197724213874955,
|
| 554 |
+
"grad_norm": 0.2525976002216339,
|
| 555 |
+
"learning_rate": 1.989541818771258e-05,
|
| 556 |
+
"loss": 0.1651,
|
| 557 |
+
"step": 670
|
| 558 |
+
},
|
| 559 |
+
{
|
| 560 |
+
"epoch": 0.08320078306619356,
|
| 561 |
+
"grad_norm": 0.25827622413635254,
|
| 562 |
+
"learning_rate": 1.9883290903825994e-05,
|
| 563 |
+
"loss": 0.1763,
|
| 564 |
+
"step": 680
|
| 565 |
+
},
|
| 566 |
+
{
|
| 567 |
+
"epoch": 0.08442432399363758,
|
| 568 |
+
"grad_norm": 0.24877820909023285,
|
| 569 |
+
"learning_rate": 1.9870502626379127e-05,
|
| 570 |
+
"loss": 0.1652,
|
| 571 |
+
"step": 690
|
| 572 |
+
},
|
| 573 |
+
{
|
| 574 |
+
"epoch": 0.0856478649210816,
|
| 575 |
+
"grad_norm": 0.278772234916687,
|
| 576 |
+
"learning_rate": 1.9857054210650776e-05,
|
| 577 |
+
"loss": 0.1651,
|
| 578 |
+
"step": 700
|
| 579 |
+
},
|
| 580 |
+
{
|
| 581 |
+
"epoch": 0.0856478649210816,
|
| 582 |
+
"eval_accuracy": 0.94905,
|
| 583 |
+
"eval_f1_weighted": 0.9490731118481598,
|
| 584 |
+
"eval_loss": 0.03884471580386162,
|
| 585 |
+
"eval_macro_f1": 0.9489981325237288,
|
| 586 |
+
"eval_precision": 0.9494795831024372,
|
| 587 |
+
"eval_recall": 0.94905,
|
| 588 |
+
"eval_runtime": 759.9337,
|
| 589 |
+
"eval_samples_per_second": 131.59,
|
| 590 |
+
"eval_steps_per_second": 1.029,
|
| 591 |
+
"step": 700
|
| 592 |
+
},
|
| 593 |
+
{
|
| 594 |
+
"epoch": 0.08687140584852564,
|
| 595 |
+
"grad_norm": 0.1895805150270462,
|
| 596 |
+
"learning_rate": 1.9842946556069705e-05,
|
| 597 |
+
"loss": 0.1628,
|
| 598 |
+
"step": 710
|
| 599 |
+
},
|
| 600 |
+
{
|
| 601 |
+
"epoch": 0.08809494677596966,
|
| 602 |
+
"grad_norm": 0.1171497106552124,
|
| 603 |
+
"learning_rate": 1.982818060615452e-05,
|
| 604 |
+
"loss": 0.1715,
|
| 605 |
+
"step": 720
|
| 606 |
+
},
|
| 607 |
+
{
|
| 608 |
+
"epoch": 0.08931848770341368,
|
| 609 |
+
"grad_norm": 0.22032728791236877,
|
| 610 |
+
"learning_rate": 1.9812757348450547e-05,
|
| 611 |
+
"loss": 0.1712,
|
| 612 |
+
"step": 730
|
| 613 |
+
},
|
| 614 |
+
{
|
| 615 |
+
"epoch": 0.0905420286308577,
|
| 616 |
+
"grad_norm": 0.20700488984584808,
|
| 617 |
+
"learning_rate": 1.9796677814463812e-05,
|
| 618 |
+
"loss": 0.1638,
|
| 619 |
+
"step": 740
|
| 620 |
+
},
|
| 621 |
+
{
|
| 622 |
+
"epoch": 0.09176556955830173,
|
| 623 |
+
"grad_norm": 0.24688613414764404,
|
| 624 |
+
"learning_rate": 1.977994307959202e-05,
|
| 625 |
+
"loss": 0.1647,
|
| 626 |
+
"step": 750
|
| 627 |
+
},
|
| 628 |
+
{
|
| 629 |
+
"epoch": 0.09298911048574575,
|
| 630 |
+
"grad_norm": 0.1886432021856308,
|
| 631 |
+
"learning_rate": 1.9762554263052655e-05,
|
| 632 |
+
"loss": 0.1734,
|
| 633 |
+
"step": 760
|
| 634 |
+
},
|
| 635 |
+
{
|
| 636 |
+
"epoch": 0.09421265141318977,
|
| 637 |
+
"grad_norm": 0.08407370001077652,
|
| 638 |
+
"learning_rate": 1.9744512527808127e-05,
|
| 639 |
+
"loss": 0.1755,
|
| 640 |
+
"step": 770
|
| 641 |
+
},
|
| 642 |
+
{
|
| 643 |
+
"epoch": 0.09543619234063379,
|
| 644 |
+
"grad_norm": 0.14160771667957306,
|
| 645 |
+
"learning_rate": 1.9725819080487976e-05,
|
| 646 |
+
"loss": 0.1749,
|
| 647 |
+
"step": 780
|
| 648 |
+
},
|
| 649 |
+
{
|
| 650 |
+
"epoch": 0.09665973326807782,
|
| 651 |
+
"grad_norm": 0.17038081586360931,
|
| 652 |
+
"learning_rate": 1.970647517130819e-05,
|
| 653 |
+
"loss": 0.1795,
|
| 654 |
+
"step": 790
|
| 655 |
+
},
|
| 656 |
+
{
|
| 657 |
+
"epoch": 0.09788327419552184,
|
| 658 |
+
"grad_norm": 0.24339532852172852,
|
| 659 |
+
"learning_rate": 1.968648209398759e-05,
|
| 660 |
+
"loss": 0.161,
|
| 661 |
+
"step": 800
|
| 662 |
+
},
|
| 663 |
+
{
|
| 664 |
+
"epoch": 0.09788327419552184,
|
| 665 |
+
"eval_accuracy": 0.95721,
|
| 666 |
+
"eval_f1_weighted": 0.9572295598206603,
|
| 667 |
+
"eval_loss": 0.03337462991476059,
|
| 668 |
+
"eval_macro_f1": 0.9571669846830866,
|
| 669 |
+
"eval_precision": 0.9576467364395861,
|
| 670 |
+
"eval_recall": 0.95721,
|
| 671 |
+
"eval_runtime": 759.9975,
|
| 672 |
+
"eval_samples_per_second": 131.579,
|
| 673 |
+
"eval_steps_per_second": 1.029,
|
| 674 |
+
"step": 800
|
| 675 |
+
},
|
| 676 |
+
{
|
| 677 |
+
"epoch": 0.09910681512296586,
|
| 678 |
+
"grad_norm": 0.16346058249473572,
|
| 679 |
+
"learning_rate": 1.9665841185661283e-05,
|
| 680 |
+
"loss": 0.1596,
|
| 681 |
+
"step": 810
|
| 682 |
+
},
|
| 683 |
+
{
|
| 684 |
+
"epoch": 0.10033035605040988,
|
| 685 |
+
"grad_norm": 0.25031423568725586,
|
| 686 |
+
"learning_rate": 1.964455382679128e-05,
|
| 687 |
+
"loss": 0.1662,
|
| 688 |
+
"step": 820
|
| 689 |
+
},
|
| 690 |
+
{
|
| 691 |
+
"epoch": 0.10155389697785391,
|
| 692 |
+
"grad_norm": 0.36678382754325867,
|
| 693 |
+
"learning_rate": 1.9622621441074113e-05,
|
| 694 |
+
"loss": 0.1552,
|
| 695 |
+
"step": 830
|
| 696 |
+
},
|
| 697 |
+
{
|
| 698 |
+
"epoch": 0.10277743790529793,
|
| 699 |
+
"grad_norm": 0.34810611605644226,
|
| 700 |
+
"learning_rate": 1.960004549534567e-05,
|
| 701 |
+
"loss": 0.1663,
|
| 702 |
+
"step": 840
|
| 703 |
+
},
|
| 704 |
+
{
|
| 705 |
+
"epoch": 0.10400097883274195,
|
| 706 |
+
"grad_norm": 0.15511688590049744,
|
| 707 |
+
"learning_rate": 1.957682749948307e-05,
|
| 708 |
+
"loss": 0.1571,
|
| 709 |
+
"step": 850
|
| 710 |
+
},
|
| 711 |
+
{
|
| 712 |
+
"epoch": 0.10522451976018597,
|
| 713 |
+
"grad_norm": 0.2082190364599228,
|
| 714 |
+
"learning_rate": 1.9552969006303666e-05,
|
| 715 |
+
"loss": 0.1497,
|
| 716 |
+
"step": 860
|
| 717 |
+
},
|
| 718 |
+
{
|
| 719 |
+
"epoch": 0.10644806068763,
|
| 720 |
+
"grad_norm": 0.14936988055706024,
|
| 721 |
+
"learning_rate": 1.9528471611461235e-05,
|
| 722 |
+
"loss": 0.1435,
|
| 723 |
+
"step": 870
|
| 724 |
+
},
|
| 725 |
+
{
|
| 726 |
+
"epoch": 0.10767160161507403,
|
| 727 |
+
"grad_norm": 0.252530574798584,
|
| 728 |
+
"learning_rate": 1.9503336953339223e-05,
|
| 729 |
+
"loss": 0.1536,
|
| 730 |
+
"step": 880
|
| 731 |
+
},
|
| 732 |
+
{
|
| 733 |
+
"epoch": 0.10889514254251804,
|
| 734 |
+
"grad_norm": 0.07253506034612656,
|
| 735 |
+
"learning_rate": 1.947756671294118e-05,
|
| 736 |
+
"loss": 0.1605,
|
| 737 |
+
"step": 890
|
| 738 |
+
},
|
| 739 |
+
{
|
| 740 |
+
"epoch": 0.11011868346996206,
|
| 741 |
+
"grad_norm": 0.09060605615377426,
|
| 742 |
+
"learning_rate": 1.9451162613778353e-05,
|
| 743 |
+
"loss": 0.1581,
|
| 744 |
+
"step": 900
|
| 745 |
+
},
|
| 746 |
+
{
|
| 747 |
+
"epoch": 0.11011868346996206,
|
| 748 |
+
"eval_accuracy": 0.95459,
|
| 749 |
+
"eval_f1_weighted": 0.9546135918699761,
|
| 750 |
+
"eval_loss": 0.03526490181684494,
|
| 751 |
+
"eval_macro_f1": 0.9545725664001194,
|
| 752 |
+
"eval_precision": 0.9559789658562307,
|
| 753 |
+
"eval_recall": 0.95459,
|
| 754 |
+
"eval_runtime": 759.7202,
|
| 755 |
+
"eval_samples_per_second": 131.627,
|
| 756 |
+
"eval_steps_per_second": 1.029,
|
| 757 |
+
"step": 900
|
| 758 |
+
},
|
| 759 |
+
{
|
| 760 |
+
"epoch": 0.1113422243974061,
|
| 761 |
+
"grad_norm": 0.21080073714256287,
|
| 762 |
+
"learning_rate": 1.942412642175439e-05,
|
| 763 |
+
"loss": 0.164,
|
| 764 |
+
"step": 910
|
| 765 |
+
},
|
| 766 |
+
{
|
| 767 |
+
"epoch": 0.11256576532485012,
|
| 768 |
+
"grad_norm": 0.22832269966602325,
|
| 769 |
+
"learning_rate": 1.939645994504726e-05,
|
| 770 |
+
"loss": 0.1595,
|
| 771 |
+
"step": 920
|
| 772 |
+
},
|
| 773 |
+
{
|
| 774 |
+
"epoch": 0.11378930625229414,
|
| 775 |
+
"grad_norm": 0.19044892489910126,
|
| 776 |
+
"learning_rate": 1.93681650339883e-05,
|
| 777 |
+
"loss": 0.1648,
|
| 778 |
+
"step": 930
|
| 779 |
+
},
|
| 780 |
+
{
|
| 781 |
+
"epoch": 0.11501284717973816,
|
| 782 |
+
"grad_norm": 0.2413518726825714,
|
| 783 |
+
"learning_rate": 1.9339243580938493e-05,
|
| 784 |
+
"loss": 0.155,
|
| 785 |
+
"step": 940
|
| 786 |
+
},
|
| 787 |
+
{
|
| 788 |
+
"epoch": 0.11623638810718219,
|
| 789 |
+
"grad_norm": 0.13218896090984344,
|
| 790 |
+
"learning_rate": 1.930969752016187e-05,
|
| 791 |
+
"loss": 0.1499,
|
| 792 |
+
"step": 950
|
| 793 |
+
},
|
| 794 |
+
{
|
| 795 |
+
"epoch": 0.11745992903462621,
|
| 796 |
+
"grad_norm": 0.23770441114902496,
|
| 797 |
+
"learning_rate": 1.927952882769619e-05,
|
| 798 |
+
"loss": 0.1555,
|
| 799 |
+
"step": 960
|
| 800 |
+
},
|
| 801 |
+
{
|
| 802 |
+
"epoch": 0.11868346996207023,
|
| 803 |
+
"grad_norm": 0.2193710207939148,
|
| 804 |
+
"learning_rate": 1.924873952122074e-05,
|
| 805 |
+
"loss": 0.1505,
|
| 806 |
+
"step": 970
|
| 807 |
+
},
|
| 808 |
+
{
|
| 809 |
+
"epoch": 0.11990701088951425,
|
| 810 |
+
"grad_norm": 0.1577174812555313,
|
| 811 |
+
"learning_rate": 1.9217331659921432e-05,
|
| 812 |
+
"loss": 0.1551,
|
| 813 |
+
"step": 980
|
| 814 |
+
},
|
| 815 |
+
{
|
| 816 |
+
"epoch": 0.12113055181695828,
|
| 817 |
+
"grad_norm": 0.1814967393875122,
|
| 818 |
+
"learning_rate": 1.9185307344353052e-05,
|
| 819 |
+
"loss": 0.1607,
|
| 820 |
+
"step": 990
|
| 821 |
+
},
|
| 822 |
+
{
|
| 823 |
+
"epoch": 0.1223540927444023,
|
| 824 |
+
"grad_norm": 0.27239707112312317,
|
| 825 |
+
"learning_rate": 1.91526687162988e-05,
|
| 826 |
+
"loss": 0.1506,
|
| 827 |
+
"step": 1000
|
| 828 |
+
},
|
| 829 |
+
{
|
| 830 |
+
"epoch": 0.1223540927444023,
|
| 831 |
+
"eval_accuracy": 0.96454,
|
| 832 |
+
"eval_f1_weighted": 0.9645510150201642,
|
| 833 |
+
"eval_loss": 0.028770776465535164,
|
| 834 |
+
"eval_macro_f1": 0.9644889377060355,
|
| 835 |
+
"eval_precision": 0.9646912114769961,
|
| 836 |
+
"eval_recall": 0.96454,
|
| 837 |
+
"eval_runtime": 759.5962,
|
| 838 |
+
"eval_samples_per_second": 131.649,
|
| 839 |
+
"eval_steps_per_second": 1.029,
|
| 840 |
+
"step": 1000
|
| 841 |
+
},
|
| 842 |
+
{
|
| 843 |
+
"epoch": 0.12357763367184632,
|
| 844 |
+
"grad_norm": 0.20511822402477264,
|
| 845 |
+
"learning_rate": 1.9119417958627022e-05,
|
| 846 |
+
"loss": 0.1503,
|
| 847 |
+
"step": 1010
|
| 848 |
+
},
|
| 849 |
+
{
|
| 850 |
+
"epoch": 0.12480117459929034,
|
| 851 |
+
"grad_norm": 0.2324151247739792,
|
| 852 |
+
"learning_rate": 1.9085557295145254e-05,
|
| 853 |
+
"loss": 0.1581,
|
| 854 |
+
"step": 1020
|
| 855 |
+
},
|
| 856 |
+
{
|
| 857 |
+
"epoch": 0.12602471552673436,
|
| 858 |
+
"grad_norm": 0.11019974201917648,
|
| 859 |
+
"learning_rate": 1.905108899045146e-05,
|
| 860 |
+
"loss": 0.1556,
|
| 861 |
+
"step": 1030
|
| 862 |
+
},
|
| 863 |
+
{
|
| 864 |
+
"epoch": 0.1272482564541784,
|
| 865 |
+
"grad_norm": 0.13370558619499207,
|
| 866 |
+
"learning_rate": 1.9016015349782594e-05,
|
| 867 |
+
"loss": 0.1469,
|
| 868 |
+
"step": 1040
|
| 869 |
+
},
|
| 870 |
+
{
|
| 871 |
+
"epoch": 0.12847179738162243,
|
| 872 |
+
"grad_norm": 0.11238931119441986,
|
| 873 |
+
"learning_rate": 1.8980338718860426e-05,
|
| 874 |
+
"loss": 0.155,
|
| 875 |
+
"step": 1050
|
| 876 |
+
},
|
| 877 |
+
{
|
| 878 |
+
"epoch": 0.12969533830906643,
|
| 879 |
+
"grad_norm": 0.2959650158882141,
|
| 880 |
+
"learning_rate": 1.894406148373466e-05,
|
| 881 |
+
"loss": 0.1478,
|
| 882 |
+
"step": 1060
|
| 883 |
+
},
|
| 884 |
+
{
|
| 885 |
+
"epoch": 0.13091887923651047,
|
| 886 |
+
"grad_norm": 0.14401015639305115,
|
| 887 |
+
"learning_rate": 1.890718607062334e-05,
|
| 888 |
+
"loss": 0.1558,
|
| 889 |
+
"step": 1070
|
| 890 |
+
},
|
| 891 |
+
{
|
| 892 |
+
"epoch": 0.13214242016395447,
|
| 893 |
+
"grad_norm": 0.2187667042016983,
|
| 894 |
+
"learning_rate": 1.8869714945750614e-05,
|
| 895 |
+
"loss": 0.1618,
|
| 896 |
+
"step": 1080
|
| 897 |
+
},
|
| 898 |
+
{
|
| 899 |
+
"epoch": 0.1333659610913985,
|
| 900 |
+
"grad_norm": 0.17045481503009796,
|
| 901 |
+
"learning_rate": 1.883165061518176e-05,
|
| 902 |
+
"loss": 0.1487,
|
| 903 |
+
"step": 1090
|
| 904 |
+
},
|
| 905 |
+
{
|
| 906 |
+
"epoch": 0.13458950201884254,
|
| 907 |
+
"grad_norm": 0.09508351981639862,
|
| 908 |
+
"learning_rate": 1.8792995624655605e-05,
|
| 909 |
+
"loss": 0.1461,
|
| 910 |
+
"step": 1100
|
| 911 |
+
},
|
| 912 |
+
{
|
| 913 |
+
"epoch": 0.13458950201884254,
|
| 914 |
+
"eval_accuracy": 0.95198,
|
| 915 |
+
"eval_f1_weighted": 0.9520023284947491,
|
| 916 |
+
"eval_loss": 0.04088817164301872,
|
| 917 |
+
"eval_macro_f1": 0.9519693367124435,
|
| 918 |
+
"eval_precision": 0.9538785423213092,
|
| 919 |
+
"eval_recall": 0.95198,
|
| 920 |
+
"eval_runtime": 759.8551,
|
| 921 |
+
"eval_samples_per_second": 131.604,
|
| 922 |
+
"eval_steps_per_second": 1.029,
|
| 923 |
+
"step": 1100
|
| 924 |
+
},
|
| 925 |
+
{
|
| 926 |
+
"epoch": 0.13581304294628654,
|
| 927 |
+
"grad_norm": 0.1880544275045395,
|
| 928 |
+
"learning_rate": 1.875375255941426e-05,
|
| 929 |
+
"loss": 0.1442,
|
| 930 |
+
"step": 1110
|
| 931 |
+
},
|
| 932 |
+
{
|
| 933 |
+
"epoch": 0.13703658387373058,
|
| 934 |
+
"grad_norm": 0.1614997684955597,
|
| 935 |
+
"learning_rate": 1.8713924044030212e-05,
|
| 936 |
+
"loss": 0.1622,
|
| 937 |
+
"step": 1120
|
| 938 |
+
},
|
| 939 |
+
{
|
| 940 |
+
"epoch": 0.1382601248011746,
|
| 941 |
+
"grad_norm": 0.10545000433921814,
|
| 942 |
+
"learning_rate": 1.86735127422308e-05,
|
| 943 |
+
"loss": 0.1531,
|
| 944 |
+
"step": 1130
|
| 945 |
+
},
|
| 946 |
+
{
|
| 947 |
+
"epoch": 0.13948366572861862,
|
| 948 |
+
"grad_norm": 0.2203921526670456,
|
| 949 |
+
"learning_rate": 1.8632521356720064e-05,
|
| 950 |
+
"loss": 0.1471,
|
| 951 |
+
"step": 1140
|
| 952 |
+
},
|
| 953 |
+
{
|
| 954 |
+
"epoch": 0.14070720665606265,
|
| 955 |
+
"grad_norm": 0.16150350868701935,
|
| 956 |
+
"learning_rate": 1.8590952628997994e-05,
|
| 957 |
+
"loss": 0.1436,
|
| 958 |
+
"step": 1150
|
| 959 |
+
},
|
| 960 |
+
{
|
| 961 |
+
"epoch": 0.14193074758350666,
|
| 962 |
+
"grad_norm": 0.1466459184885025,
|
| 963 |
+
"learning_rate": 1.8548809339177155e-05,
|
| 964 |
+
"loss": 0.1486,
|
| 965 |
+
"step": 1160
|
| 966 |
+
},
|
| 967 |
+
{
|
| 968 |
+
"epoch": 0.1431542885109507,
|
| 969 |
+
"grad_norm": 0.1858566850423813,
|
| 970 |
+
"learning_rate": 1.8506094305796786e-05,
|
| 971 |
+
"loss": 0.1485,
|
| 972 |
+
"step": 1170
|
| 973 |
+
},
|
| 974 |
+
{
|
| 975 |
+
"epoch": 0.14437782943839472,
|
| 976 |
+
"grad_norm": 0.21629975736141205,
|
| 977 |
+
"learning_rate": 1.846281038563428e-05,
|
| 978 |
+
"loss": 0.1558,
|
| 979 |
+
"step": 1180
|
| 980 |
+
},
|
| 981 |
+
{
|
| 982 |
+
"epoch": 0.14560137036583873,
|
| 983 |
+
"grad_norm": 0.12939126789569855,
|
| 984 |
+
"learning_rate": 1.841896047351412e-05,
|
| 985 |
+
"loss": 0.1388,
|
| 986 |
+
"step": 1190
|
| 987 |
+
},
|
| 988 |
+
{
|
| 989 |
+
"epoch": 0.14682491129328276,
|
| 990 |
+
"grad_norm": 0.12447889149188995,
|
| 991 |
+
"learning_rate": 1.8374547502114277e-05,
|
| 992 |
+
"loss": 0.1683,
|
| 993 |
+
"step": 1200
|
| 994 |
+
},
|
| 995 |
+
{
|
| 996 |
+
"epoch": 0.14682491129328276,
|
| 997 |
+
"eval_accuracy": 0.96627,
|
| 998 |
+
"eval_f1_weighted": 0.9662808443914441,
|
| 999 |
+
"eval_loss": 0.027206970378756523,
|
| 1000 |
+
"eval_macro_f1": 0.9662223734621376,
|
| 1001 |
+
"eval_precision": 0.966433068847901,
|
| 1002 |
+
"eval_recall": 0.96627,
|
| 1003 |
+
"eval_runtime": 661.8465,
|
| 1004 |
+
"eval_samples_per_second": 151.092,
|
| 1005 |
+
"eval_steps_per_second": 1.182,
|
| 1006 |
+
"step": 1200
|
| 1007 |
+
},
|
| 1008 |
+
{
|
| 1009 |
+
"epoch": 0.1480484522207268,
|
| 1010 |
+
"grad_norm": 0.07575482875108719,
|
| 1011 |
+
"learning_rate": 1.832957444177007e-05,
|
| 1012 |
+
"loss": 0.152,
|
| 1013 |
+
"step": 1210
|
| 1014 |
+
},
|
| 1015 |
+
{
|
| 1016 |
+
"epoch": 0.1492719931481708,
|
| 1017 |
+
"grad_norm": 0.22879591584205627,
|
| 1018 |
+
"learning_rate": 1.828404430027553e-05,
|
| 1019 |
+
"loss": 0.15,
|
| 1020 |
+
"step": 1220
|
| 1021 |
+
},
|
| 1022 |
+
{
|
| 1023 |
+
"epoch": 0.15049553407561483,
|
| 1024 |
+
"grad_norm": 0.25837475061416626,
|
| 1025 |
+
"learning_rate": 1.8237960122682207e-05,
|
| 1026 |
+
"loss": 0.1557,
|
| 1027 |
+
"step": 1230
|
| 1028 |
+
},
|
| 1029 |
+
{
|
| 1030 |
+
"epoch": 0.15171907500305884,
|
| 1031 |
+
"grad_norm": 0.1440538763999939,
|
| 1032 |
+
"learning_rate": 1.8191324991095527e-05,
|
| 1033 |
+
"loss": 0.154,
|
| 1034 |
+
"step": 1240
|
| 1035 |
+
},
|
| 1036 |
+
{
|
| 1037 |
+
"epoch": 0.15294261593050287,
|
| 1038 |
+
"grad_norm": 0.21304763853549957,
|
| 1039 |
+
"learning_rate": 1.814414202446869e-05,
|
| 1040 |
+
"loss": 0.1428,
|
| 1041 |
+
"step": 1250
|
| 1042 |
+
},
|
| 1043 |
+
{
|
| 1044 |
+
"epoch": 0.1541661568579469,
|
| 1045 |
+
"grad_norm": 0.2292693555355072,
|
| 1046 |
+
"learning_rate": 1.809641437839403e-05,
|
| 1047 |
+
"loss": 0.1441,
|
| 1048 |
+
"step": 1260
|
| 1049 |
+
},
|
| 1050 |
+
{
|
| 1051 |
+
"epoch": 0.1553896977853909,
|
| 1052 |
+
"grad_norm": 0.0891769602894783,
|
| 1053 |
+
"learning_rate": 1.8048145244892e-05,
|
| 1054 |
+
"loss": 0.1563,
|
| 1055 |
+
"step": 1270
|
| 1056 |
+
},
|
| 1057 |
+
{
|
| 1058 |
+
"epoch": 0.15661323871283495,
|
| 1059 |
+
"grad_norm": 0.13292840123176575,
|
| 1060 |
+
"learning_rate": 1.799933785219769e-05,
|
| 1061 |
+
"loss": 0.1418,
|
| 1062 |
+
"step": 1280
|
| 1063 |
+
},
|
| 1064 |
+
{
|
| 1065 |
+
"epoch": 0.15783677964027898,
|
| 1066 |
+
"grad_norm": 0.30031099915504456,
|
| 1067 |
+
"learning_rate": 1.7949995464544892e-05,
|
| 1068 |
+
"loss": 0.1543,
|
| 1069 |
+
"step": 1290
|
| 1070 |
+
},
|
| 1071 |
+
{
|
| 1072 |
+
"epoch": 0.15906032056772298,
|
| 1073 |
+
"grad_norm": 0.08812980353832245,
|
| 1074 |
+
"learning_rate": 1.790012138194783e-05,
|
| 1075 |
+
"loss": 0.1473,
|
| 1076 |
+
"step": 1300
|
| 1077 |
+
},
|
| 1078 |
+
{
|
| 1079 |
+
"epoch": 0.15906032056772298,
|
| 1080 |
+
"eval_accuracy": 0.96636,
|
| 1081 |
+
"eval_f1_weighted": 0.9663763998373325,
|
| 1082 |
+
"eval_loss": 0.02714269980788231,
|
| 1083 |
+
"eval_macro_f1": 0.9663303002945869,
|
| 1084 |
+
"eval_precision": 0.9669126928794434,
|
| 1085 |
+
"eval_recall": 0.96636,
|
| 1086 |
+
"eval_runtime": 661.9303,
|
| 1087 |
+
"eval_samples_per_second": 151.073,
|
| 1088 |
+
"eval_steps_per_second": 1.181,
|
| 1089 |
+
"step": 1300
|
| 1090 |
+
},
|
| 1091 |
+
{
|
| 1092 |
+
"epoch": 0.16028386149516702,
|
| 1093 |
+
"grad_norm": 0.1380453258752823,
|
| 1094 |
+
"learning_rate": 1.7849718939980426e-05,
|
| 1095 |
+
"loss": 0.1594,
|
| 1096 |
+
"step": 1310
|
| 1097 |
+
},
|
| 1098 |
+
{
|
| 1099 |
+
"epoch": 0.16150740242261102,
|
| 1100 |
+
"grad_norm": 0.08388533443212509,
|
| 1101 |
+
"learning_rate": 1.7798791509553224e-05,
|
| 1102 |
+
"loss": 0.1411,
|
| 1103 |
+
"step": 1320
|
| 1104 |
+
},
|
| 1105 |
+
{
|
| 1106 |
+
"epoch": 0.16273094335005506,
|
| 1107 |
+
"grad_norm": 0.2659275531768799,
|
| 1108 |
+
"learning_rate": 1.7747342496687957e-05,
|
| 1109 |
+
"loss": 0.1495,
|
| 1110 |
+
"step": 1330
|
| 1111 |
+
},
|
| 1112 |
+
{
|
| 1113 |
+
"epoch": 0.1639544842774991,
|
| 1114 |
+
"grad_norm": 0.275398313999176,
|
| 1115 |
+
"learning_rate": 1.7695375342289732e-05,
|
| 1116 |
+
"loss": 0.134,
|
| 1117 |
+
"step": 1340
|
| 1118 |
+
},
|
| 1119 |
+
{
|
| 1120 |
+
"epoch": 0.1651780252049431,
|
| 1121 |
+
"grad_norm": 0.15272289514541626,
|
| 1122 |
+
"learning_rate": 1.7642893521916926e-05,
|
| 1123 |
+
"loss": 0.1427,
|
| 1124 |
+
"step": 1350
|
| 1125 |
+
},
|
| 1126 |
+
{
|
| 1127 |
+
"epoch": 0.16640156613238713,
|
| 1128 |
+
"grad_norm": 0.21478359401226044,
|
| 1129 |
+
"learning_rate": 1.7589900545548706e-05,
|
| 1130 |
+
"loss": 0.1465,
|
| 1131 |
+
"step": 1360
|
| 1132 |
+
},
|
| 1133 |
+
{
|
| 1134 |
+
"epoch": 0.16762510705983116,
|
| 1135 |
+
"grad_norm": 0.2184278517961502,
|
| 1136 |
+
"learning_rate": 1.7536399957350326e-05,
|
| 1137 |
+
"loss": 0.1452,
|
| 1138 |
+
"step": 1370
|
| 1139 |
+
},
|
| 1140 |
+
{
|
| 1141 |
+
"epoch": 0.16884864798727517,
|
| 1142 |
+
"grad_norm": 0.23272652924060822,
|
| 1143 |
+
"learning_rate": 1.7482395335436052e-05,
|
| 1144 |
+
"loss": 0.142,
|
| 1145 |
+
"step": 1380
|
| 1146 |
+
},
|
| 1147 |
+
{
|
| 1148 |
+
"epoch": 0.1700721889147192,
|
| 1149 |
+
"grad_norm": 0.14892522990703583,
|
| 1150 |
+
"learning_rate": 1.7427890291629895e-05,
|
| 1151 |
+
"loss": 0.1465,
|
| 1152 |
+
"step": 1390
|
| 1153 |
+
},
|
| 1154 |
+
{
|
| 1155 |
+
"epoch": 0.1712957298421632,
|
| 1156 |
+
"grad_norm": 0.2066282033920288,
|
| 1157 |
+
"learning_rate": 1.737288847122402e-05,
|
| 1158 |
+
"loss": 0.1444,
|
| 1159 |
+
"step": 1400
|
| 1160 |
+
},
|
| 1161 |
+
{
|
| 1162 |
+
"epoch": 0.1712957298421632,
|
| 1163 |
+
"eval_accuracy": 0.95302,
|
| 1164 |
+
"eval_f1_weighted": 0.9530351521706546,
|
| 1165 |
+
"eval_loss": 0.03864819556474686,
|
| 1166 |
+
"eval_macro_f1": 0.9530165251021966,
|
| 1167 |
+
"eval_precision": 0.9557329367218795,
|
| 1168 |
+
"eval_recall": 0.95302,
|
| 1169 |
+
"eval_runtime": 661.7384,
|
| 1170 |
+
"eval_samples_per_second": 151.117,
|
| 1171 |
+
"eval_steps_per_second": 1.182,
|
| 1172 |
+
"step": 1400
|
| 1173 |
+
},
|
| 1174 |
+
{
|
| 1175 |
+
"epoch": 0.17251927076960724,
|
| 1176 |
+
"grad_norm": 0.2941270172595978,
|
| 1177 |
+
"learning_rate": 1.7317393552734977e-05,
|
| 1178 |
+
"loss": 0.136,
|
| 1179 |
+
"step": 1410
|
| 1180 |
+
},
|
| 1181 |
+
{
|
| 1182 |
+
"epoch": 0.17374281169705127,
|
| 1183 |
+
"grad_norm": 0.11988822370767593,
|
| 1184 |
+
"learning_rate": 1.7261409247657656e-05,
|
| 1185 |
+
"loss": 0.1551,
|
| 1186 |
+
"step": 1420
|
| 1187 |
+
},
|
| 1188 |
+
{
|
| 1189 |
+
"epoch": 0.17496635262449528,
|
| 1190 |
+
"grad_norm": 0.346236914396286,
|
| 1191 |
+
"learning_rate": 1.7204939300217084e-05,
|
| 1192 |
+
"loss": 0.1453,
|
| 1193 |
+
"step": 1430
|
| 1194 |
+
},
|
| 1195 |
+
{
|
| 1196 |
+
"epoch": 0.17618989355193931,
|
| 1197 |
+
"grad_norm": 0.15973429381847382,
|
| 1198 |
+
"learning_rate": 1.7147987487118006e-05,
|
| 1199 |
+
"loss": 0.1341,
|
| 1200 |
+
"step": 1440
|
| 1201 |
+
},
|
| 1202 |
+
{
|
| 1203 |
+
"epoch": 0.17741343447938335,
|
| 1204 |
+
"grad_norm": 0.2531983554363251,
|
| 1205 |
+
"learning_rate": 1.7090557617292287e-05,
|
| 1206 |
+
"loss": 0.1479,
|
| 1207 |
+
"step": 1450
|
| 1208 |
+
},
|
| 1209 |
+
{
|
| 1210 |
+
"epoch": 0.17863697540682735,
|
| 1211 |
+
"grad_norm": 0.10105066746473312,
|
| 1212 |
+
"learning_rate": 1.703265353164419e-05,
|
| 1213 |
+
"loss": 0.1459,
|
| 1214 |
+
"step": 1460
|
| 1215 |
+
},
|
| 1216 |
+
{
|
| 1217 |
+
"epoch": 0.1798605163342714,
|
| 1218 |
+
"grad_norm": 0.10651528090238571,
|
| 1219 |
+
"learning_rate": 1.6974279102793473e-05,
|
| 1220 |
+
"loss": 0.152,
|
| 1221 |
+
"step": 1470
|
| 1222 |
+
},
|
| 1223 |
+
{
|
| 1224 |
+
"epoch": 0.1810840572617154,
|
| 1225 |
+
"grad_norm": 0.2675159275531769,
|
| 1226 |
+
"learning_rate": 1.6915438234816423e-05,
|
| 1227 |
+
"loss": 0.1467,
|
| 1228 |
+
"step": 1480
|
| 1229 |
+
},
|
| 1230 |
+
{
|
| 1231 |
+
"epoch": 0.18230759818915943,
|
| 1232 |
+
"grad_norm": 0.09507659822702408,
|
| 1233 |
+
"learning_rate": 1.6856134862984718e-05,
|
| 1234 |
+
"loss": 0.1504,
|
| 1235 |
+
"step": 1490
|
| 1236 |
+
},
|
| 1237 |
+
{
|
| 1238 |
+
"epoch": 0.18353113911660346,
|
| 1239 |
+
"grad_norm": 0.13831426203250885,
|
| 1240 |
+
"learning_rate": 1.6796372953502252e-05,
|
| 1241 |
+
"loss": 0.151,
|
| 1242 |
+
"step": 1500
|
| 1243 |
+
},
|
| 1244 |
+
{
|
| 1245 |
+
"epoch": 0.18353113911660346,
|
| 1246 |
+
"eval_accuracy": 0.96716,
|
| 1247 |
+
"eval_f1_weighted": 0.9671757267424962,
|
| 1248 |
+
"eval_loss": 0.02643330581486225,
|
| 1249 |
+
"eval_macro_f1": 0.9671297823300075,
|
| 1250 |
+
"eval_precision": 0.9676731949933907,
|
| 1251 |
+
"eval_recall": 0.96716,
|
| 1252 |
+
"eval_runtime": 662.2285,
|
| 1253 |
+
"eval_samples_per_second": 151.005,
|
| 1254 |
+
"eval_steps_per_second": 1.181,
|
| 1255 |
+
"step": 1500
|
| 1256 |
+
},
|
| 1257 |
+
{
|
| 1258 |
+
"epoch": 0.18475468004404746,
|
| 1259 |
+
"grad_norm": 0.20989136397838593,
|
| 1260 |
+
"learning_rate": 1.673615650323988e-05,
|
| 1261 |
+
"loss": 0.1424,
|
| 1262 |
+
"step": 1510
|
| 1263 |
+
},
|
| 1264 |
+
{
|
| 1265 |
+
"epoch": 0.1859782209714915,
|
| 1266 |
+
"grad_norm": 0.13308484852313995,
|
| 1267 |
+
"learning_rate": 1.6675489539468094e-05,
|
| 1268 |
+
"loss": 0.1541,
|
| 1269 |
+
"step": 1520
|
| 1270 |
+
},
|
| 1271 |
+
{
|
| 1272 |
+
"epoch": 0.18720176189893553,
|
| 1273 |
+
"grad_norm": 0.3802664577960968,
|
| 1274 |
+
"learning_rate": 1.6614376119587685e-05,
|
| 1275 |
+
"loss": 0.1426,
|
| 1276 |
+
"step": 1530
|
| 1277 |
+
},
|
| 1278 |
+
{
|
| 1279 |
+
"epoch": 0.18842530282637954,
|
| 1280 |
+
"grad_norm": 0.254869282245636,
|
| 1281 |
+
"learning_rate": 1.6552820330858398e-05,
|
| 1282 |
+
"loss": 0.1407,
|
| 1283 |
+
"step": 1540
|
| 1284 |
+
},
|
| 1285 |
+
{
|
| 1286 |
+
"epoch": 0.18964884375382357,
|
| 1287 |
+
"grad_norm": 0.22702300548553467,
|
| 1288 |
+
"learning_rate": 1.6490826290125555e-05,
|
| 1289 |
+
"loss": 0.1404,
|
| 1290 |
+
"step": 1550
|
| 1291 |
+
},
|
| 1292 |
+
{
|
| 1293 |
+
"epoch": 0.19087238468126758,
|
| 1294 |
+
"grad_norm": 0.18147043883800507,
|
| 1295 |
+
"learning_rate": 1.6428398143544727e-05,
|
| 1296 |
+
"loss": 0.1398,
|
| 1297 |
+
"step": 1560
|
| 1298 |
+
},
|
| 1299 |
+
{
|
| 1300 |
+
"epoch": 0.1920959256087116,
|
| 1301 |
+
"grad_norm": 0.2697111964225769,
|
| 1302 |
+
"learning_rate": 1.636554006630446e-05,
|
| 1303 |
+
"loss": 0.1465,
|
| 1304 |
+
"step": 1570
|
| 1305 |
+
},
|
| 1306 |
+
{
|
| 1307 |
+
"epoch": 0.19331946653615564,
|
| 1308 |
+
"grad_norm": 0.22285781800746918,
|
| 1309 |
+
"learning_rate": 1.6302256262347e-05,
|
| 1310 |
+
"loss": 0.1398,
|
| 1311 |
+
"step": 1580
|
| 1312 |
+
},
|
| 1313 |
+
{
|
| 1314 |
+
"epoch": 0.19454300746359965,
|
| 1315 |
+
"grad_norm": 0.2551407814025879,
|
| 1316 |
+
"learning_rate": 1.6238550964087175e-05,
|
| 1317 |
+
"loss": 0.1347,
|
| 1318 |
+
"step": 1590
|
| 1319 |
+
},
|
| 1320 |
+
{
|
| 1321 |
+
"epoch": 0.19576654839104368,
|
| 1322 |
+
"grad_norm": 0.25052276253700256,
|
| 1323 |
+
"learning_rate": 1.61744284321293e-05,
|
| 1324 |
+
"loss": 0.1461,
|
| 1325 |
+
"step": 1600
|
| 1326 |
+
},
|
| 1327 |
+
{
|
| 1328 |
+
"epoch": 0.19576654839104368,
|
| 1329 |
+
"eval_accuracy": 0.96526,
|
| 1330 |
+
"eval_f1_weighted": 0.9652784309636304,
|
| 1331 |
+
"eval_loss": 0.02899531088769436,
|
| 1332 |
+
"eval_macro_f1": 0.9652431164406312,
|
| 1333 |
+
"eval_precision": 0.9664143950990177,
|
| 1334 |
+
"eval_recall": 0.96526,
|
| 1335 |
+
"eval_runtime": 662.3587,
|
| 1336 |
+
"eval_samples_per_second": 150.976,
|
| 1337 |
+
"eval_steps_per_second": 1.181,
|
| 1338 |
+
"step": 1600
|
| 1339 |
+
},
|
| 1340 |
+
{
|
| 1341 |
+
"epoch": 0.19699008931848772,
|
| 1342 |
+
"grad_norm": 0.09207922220230103,
|
| 1343 |
+
"learning_rate": 1.6109892954982242e-05,
|
| 1344 |
+
"loss": 0.1382,
|
| 1345 |
+
"step": 1610
|
| 1346 |
+
},
|
| 1347 |
+
{
|
| 1348 |
+
"epoch": 0.19821363024593172,
|
| 1349 |
+
"grad_norm": 0.2670917809009552,
|
| 1350 |
+
"learning_rate": 1.6044948848772605e-05,
|
| 1351 |
+
"loss": 0.142,
|
| 1352 |
+
"step": 1620
|
| 1353 |
+
},
|
| 1354 |
+
{
|
| 1355 |
+
"epoch": 0.19943717117337575,
|
| 1356 |
+
"grad_norm": 0.28024110198020935,
|
| 1357 |
+
"learning_rate": 1.597960045695607e-05,
|
| 1358 |
+
"loss": 0.1385,
|
| 1359 |
+
"step": 1630
|
| 1360 |
+
},
|
| 1361 |
+
{
|
| 1362 |
+
"epoch": 0.20066071210081976,
|
| 1363 |
+
"grad_norm": 0.15802879631519318,
|
| 1364 |
+
"learning_rate": 1.5913852150026895e-05,
|
| 1365 |
+
"loss": 0.1328,
|
| 1366 |
+
"step": 1640
|
| 1367 |
+
},
|
| 1368 |
+
{
|
| 1369 |
+
"epoch": 0.2018842530282638,
|
| 1370 |
+
"grad_norm": 0.12178096175193787,
|
| 1371 |
+
"learning_rate": 1.5847708325225618e-05,
|
| 1372 |
+
"loss": 0.1446,
|
| 1373 |
+
"step": 1650
|
| 1374 |
+
},
|
| 1375 |
+
{
|
| 1376 |
+
"epoch": 0.20310779395570783,
|
| 1377 |
+
"grad_norm": 0.22587499022483826,
|
| 1378 |
+
"learning_rate": 1.578117340624499e-05,
|
| 1379 |
+
"loss": 0.1406,
|
| 1380 |
+
"step": 1660
|
| 1381 |
+
},
|
| 1382 |
+
{
|
| 1383 |
+
"epoch": 0.20433133488315183,
|
| 1384 |
+
"grad_norm": 0.18149550259113312,
|
| 1385 |
+
"learning_rate": 1.5714251842934096e-05,
|
| 1386 |
+
"loss": 0.1459,
|
| 1387 |
+
"step": 1670
|
| 1388 |
+
},
|
| 1389 |
+
{
|
| 1390 |
+
"epoch": 0.20555487581059587,
|
| 1391 |
+
"grad_norm": 0.15662655234336853,
|
| 1392 |
+
"learning_rate": 1.5646948111000756e-05,
|
| 1393 |
+
"loss": 0.1344,
|
| 1394 |
+
"step": 1680
|
| 1395 |
+
},
|
| 1396 |
+
{
|
| 1397 |
+
"epoch": 0.2067784167380399,
|
| 1398 |
+
"grad_norm": 0.1990954577922821,
|
| 1399 |
+
"learning_rate": 1.5579266711712185e-05,
|
| 1400 |
+
"loss": 0.1384,
|
| 1401 |
+
"step": 1690
|
| 1402 |
+
},
|
| 1403 |
+
{
|
| 1404 |
+
"epoch": 0.2080019576654839,
|
| 1405 |
+
"grad_norm": 0.14277924597263336,
|
| 1406 |
+
"learning_rate": 1.551121217159397e-05,
|
| 1407 |
+
"loss": 0.1251,
|
| 1408 |
+
"step": 1700
|
| 1409 |
+
},
|
| 1410 |
+
{
|
| 1411 |
+
"epoch": 0.2080019576654839,
|
| 1412 |
+
"eval_accuracy": 0.97199,
|
| 1413 |
+
"eval_f1_weighted": 0.9719965673787706,
|
| 1414 |
+
"eval_loss": 0.025022055953741074,
|
| 1415 |
+
"eval_macro_f1": 0.9719444578845672,
|
| 1416 |
+
"eval_precision": 0.9720688283360513,
|
| 1417 |
+
"eval_recall": 0.97199,
|
| 1418 |
+
"eval_runtime": 661.2104,
|
| 1419 |
+
"eval_samples_per_second": 151.238,
|
| 1420 |
+
"eval_steps_per_second": 1.183,
|
| 1421 |
+
"step": 1700
|
| 1422 |
+
},
|
| 1423 |
+
{
|
| 1424 |
+
"epoch": 0.20922549859292794,
|
| 1425 |
+
"grad_norm": 0.13305141031742096,
|
| 1426 |
+
"learning_rate": 1.5442789042127306e-05,
|
| 1427 |
+
"loss": 0.1465,
|
| 1428 |
+
"step": 1710
|
| 1429 |
+
},
|
| 1430 |
+
{
|
| 1431 |
+
"epoch": 0.21044903952037194,
|
| 1432 |
+
"grad_norm": 0.2761312425136566,
|
| 1433 |
+
"learning_rate": 1.5374001899444624e-05,
|
| 1434 |
+
"loss": 0.1413,
|
| 1435 |
+
"step": 1720
|
| 1436 |
+
},
|
| 1437 |
+
{
|
| 1438 |
+
"epoch": 0.21167258044781598,
|
| 1439 |
+
"grad_norm": 0.11722280085086823,
|
| 1440 |
+
"learning_rate": 1.530485534402351e-05,
|
| 1441 |
+
"loss": 0.1337,
|
| 1442 |
+
"step": 1730
|
| 1443 |
+
},
|
| 1444 |
+
{
|
| 1445 |
+
"epoch": 0.21289612137526,
|
| 1446 |
+
"grad_norm": 0.0893188863992691,
|
| 1447 |
+
"learning_rate": 1.523535400037905e-05,
|
| 1448 |
+
"loss": 0.1458,
|
| 1449 |
+
"step": 1740
|
| 1450 |
+
},
|
| 1451 |
+
{
|
| 1452 |
+
"epoch": 0.21411966230270402,
|
| 1453 |
+
"grad_norm": 0.14586731791496277,
|
| 1454 |
+
"learning_rate": 1.5165502516754526e-05,
|
| 1455 |
+
"loss": 0.1389,
|
| 1456 |
+
"step": 1750
|
| 1457 |
+
},
|
| 1458 |
+
{
|
| 1459 |
+
"epoch": 0.21534320323014805,
|
| 1460 |
+
"grad_norm": 0.23379822075366974,
|
| 1461 |
+
"learning_rate": 1.509530556481056e-05,
|
| 1462 |
+
"loss": 0.1467,
|
| 1463 |
+
"step": 1760
|
| 1464 |
+
},
|
| 1465 |
+
{
|
| 1466 |
+
"epoch": 0.21656674415759208,
|
| 1467 |
+
"grad_norm": 0.12035554647445679,
|
| 1468 |
+
"learning_rate": 1.5024767839312642e-05,
|
| 1469 |
+
"loss": 0.1386,
|
| 1470 |
+
"step": 1770
|
| 1471 |
+
},
|
| 1472 |
+
{
|
| 1473 |
+
"epoch": 0.2177902850850361,
|
| 1474 |
+
"grad_norm": 0.1451006382703781,
|
| 1475 |
+
"learning_rate": 1.495389405781719e-05,
|
| 1476 |
+
"loss": 0.1435,
|
| 1477 |
+
"step": 1780
|
| 1478 |
+
},
|
| 1479 |
+
{
|
| 1480 |
+
"epoch": 0.21901382601248012,
|
| 1481 |
+
"grad_norm": 0.08014333248138428,
|
| 1482 |
+
"learning_rate": 1.4882688960355994e-05,
|
| 1483 |
+
"loss": 0.1491,
|
| 1484 |
+
"step": 1790
|
| 1485 |
+
},
|
| 1486 |
+
{
|
| 1487 |
+
"epoch": 0.22023736693992413,
|
| 1488 |
+
"grad_norm": 0.17279869318008423,
|
| 1489 |
+
"learning_rate": 1.4811157309119235e-05,
|
| 1490 |
+
"loss": 0.151,
|
| 1491 |
+
"step": 1800
|
| 1492 |
+
},
|
| 1493 |
+
{
|
| 1494 |
+
"epoch": 0.22023736693992413,
|
| 1495 |
+
"eval_accuracy": 0.96674,
|
| 1496 |
+
"eval_f1_weighted": 0.9667576374640046,
|
| 1497 |
+
"eval_loss": 0.02695448324084282,
|
| 1498 |
+
"eval_macro_f1": 0.9667205101995933,
|
| 1499 |
+
"eval_precision": 0.9677025757044254,
|
| 1500 |
+
"eval_recall": 0.96674,
|
| 1501 |
+
"eval_runtime": 661.5931,
|
| 1502 |
+
"eval_samples_per_second": 151.15,
|
| 1503 |
+
"eval_steps_per_second": 1.182,
|
| 1504 |
+
"step": 1800
|
| 1505 |
+
},
|
| 1506 |
+
{
|
| 1507 |
+
"epoch": 0.22146090786736816,
|
| 1508 |
+
"grad_norm": 0.32028236985206604,
|
| 1509 |
+
"learning_rate": 1.473930388813698e-05,
|
| 1510 |
+
"loss": 0.1342,
|
| 1511 |
+
"step": 1810
|
| 1512 |
+
},
|
| 1513 |
+
{
|
| 1514 |
+
"epoch": 0.2226844487948122,
|
| 1515 |
+
"grad_norm": 0.2838447690010071,
|
| 1516 |
+
"learning_rate": 1.466713350295922e-05,
|
| 1517 |
+
"loss": 0.1346,
|
| 1518 |
+
"step": 1820
|
| 1519 |
+
},
|
| 1520 |
+
{
|
| 1521 |
+
"epoch": 0.2239079897222562,
|
| 1522 |
+
"grad_norm": 0.2165336161851883,
|
| 1523 |
+
"learning_rate": 1.4594650980334492e-05,
|
| 1524 |
+
"loss": 0.1447,
|
| 1525 |
+
"step": 1830
|
| 1526 |
+
},
|
| 1527 |
+
{
|
| 1528 |
+
"epoch": 0.22513153064970023,
|
| 1529 |
+
"grad_norm": 0.1966007500886917,
|
| 1530 |
+
"learning_rate": 1.4521861167887043e-05,
|
| 1531 |
+
"loss": 0.1405,
|
| 1532 |
+
"step": 1840
|
| 1533 |
+
},
|
| 1534 |
+
{
|
| 1535 |
+
"epoch": 0.22635507157714427,
|
| 1536 |
+
"grad_norm": 0.16725124418735504,
|
| 1537 |
+
"learning_rate": 1.4448768933792644e-05,
|
| 1538 |
+
"loss": 0.1272,
|
| 1539 |
+
"step": 1850
|
| 1540 |
+
},
|
| 1541 |
+
{
|
| 1542 |
+
"epoch": 0.22757861250458827,
|
| 1543 |
+
"grad_norm": 0.22608086466789246,
|
| 1544 |
+
"learning_rate": 1.4375379166452991e-05,
|
| 1545 |
+
"loss": 0.1305,
|
| 1546 |
+
"step": 1860
|
| 1547 |
+
},
|
| 1548 |
+
{
|
| 1549 |
+
"epoch": 0.2288021534320323,
|
| 1550 |
+
"grad_norm": 0.24929635226726532,
|
| 1551 |
+
"learning_rate": 1.4301696774168782e-05,
|
| 1552 |
+
"loss": 0.1276,
|
| 1553 |
+
"step": 1870
|
| 1554 |
+
},
|
| 1555 |
+
{
|
| 1556 |
+
"epoch": 0.2300256943594763,
|
| 1557 |
+
"grad_norm": 0.0894479975104332,
|
| 1558 |
+
"learning_rate": 1.4227726684811439e-05,
|
| 1559 |
+
"loss": 0.1515,
|
| 1560 |
+
"step": 1880
|
| 1561 |
+
},
|
| 1562 |
+
{
|
| 1563 |
+
"epoch": 0.23124923528692035,
|
| 1564 |
+
"grad_norm": 0.31293338537216187,
|
| 1565 |
+
"learning_rate": 1.4153473845493531e-05,
|
| 1566 |
+
"loss": 0.1262,
|
| 1567 |
+
"step": 1890
|
| 1568 |
+
},
|
| 1569 |
+
{
|
| 1570 |
+
"epoch": 0.23247277621436438,
|
| 1571 |
+
"grad_norm": 0.34327465295791626,
|
| 1572 |
+
"learning_rate": 1.4078943222237933e-05,
|
| 1573 |
+
"loss": 0.1371,
|
| 1574 |
+
"step": 1900
|
| 1575 |
+
},
|
| 1576 |
+
{
|
| 1577 |
+
"epoch": 0.23247277621436438,
|
| 1578 |
+
"eval_accuracy": 0.97149,
|
| 1579 |
+
"eval_f1_weighted": 0.9715031773882226,
|
| 1580 |
+
"eval_loss": 0.02571570873260498,
|
| 1581 |
+
"eval_macro_f1": 0.9714618804747581,
|
| 1582 |
+
"eval_precision": 0.9719368831197501,
|
| 1583 |
+
"eval_recall": 0.97149,
|
| 1584 |
+
"eval_runtime": 660.9647,
|
| 1585 |
+
"eval_samples_per_second": 151.294,
|
| 1586 |
+
"eval_steps_per_second": 1.183,
|
| 1587 |
+
"step": 1900
|
| 1588 |
+
},
|
| 1589 |
+
{
|
| 1590 |
+
"epoch": 0.23369631714180839,
|
| 1591 |
+
"grad_norm": 0.07504434138536453,
|
| 1592 |
+
"learning_rate": 1.4004139799645669e-05,
|
| 1593 |
+
"loss": 0.1506,
|
| 1594 |
+
"step": 1910
|
| 1595 |
+
},
|
| 1596 |
+
{
|
| 1597 |
+
"epoch": 0.23491985806925242,
|
| 1598 |
+
"grad_norm": 0.2954120934009552,
|
| 1599 |
+
"learning_rate": 1.3929068580562561e-05,
|
| 1600 |
+
"loss": 0.1329,
|
| 1601 |
+
"step": 1920
|
| 1602 |
+
},
|
| 1603 |
+
{
|
| 1604 |
+
"epoch": 0.23614339899669645,
|
| 1605 |
+
"grad_norm": 0.17359323799610138,
|
| 1606 |
+
"learning_rate": 1.3853734585744642e-05,
|
| 1607 |
+
"loss": 0.1347,
|
| 1608 |
+
"step": 1930
|
| 1609 |
+
},
|
| 1610 |
+
{
|
| 1611 |
+
"epoch": 0.23736693992414046,
|
| 1612 |
+
"grad_norm": 0.15386463701725006,
|
| 1613 |
+
"learning_rate": 1.3778142853522352e-05,
|
| 1614 |
+
"loss": 0.1429,
|
| 1615 |
+
"step": 1940
|
| 1616 |
+
},
|
| 1617 |
+
{
|
| 1618 |
+
"epoch": 0.2385904808515845,
|
| 1619 |
+
"grad_norm": 0.2623615264892578,
|
| 1620 |
+
"learning_rate": 1.3702298439463589e-05,
|
| 1621 |
+
"loss": 0.1296,
|
| 1622 |
+
"step": 1950
|
| 1623 |
+
},
|
| 1624 |
+
{
|
| 1625 |
+
"epoch": 0.2398140217790285,
|
| 1626 |
+
"grad_norm": 0.12422071397304535,
|
| 1627 |
+
"learning_rate": 1.3626206416035589e-05,
|
| 1628 |
+
"loss": 0.1441,
|
| 1629 |
+
"step": 1960
|
| 1630 |
+
},
|
| 1631 |
+
{
|
| 1632 |
+
"epoch": 0.24103756270647253,
|
| 1633 |
+
"grad_norm": 0.14957550168037415,
|
| 1634 |
+
"learning_rate": 1.3549871872265672e-05,
|
| 1635 |
+
"loss": 0.1406,
|
| 1636 |
+
"step": 1970
|
| 1637 |
+
},
|
| 1638 |
+
{
|
| 1639 |
+
"epoch": 0.24226110363391656,
|
| 1640 |
+
"grad_norm": 0.11441356688737869,
|
| 1641 |
+
"learning_rate": 1.3473299913400905e-05,
|
| 1642 |
+
"loss": 0.1294,
|
| 1643 |
+
"step": 1980
|
| 1644 |
+
},
|
| 1645 |
+
{
|
| 1646 |
+
"epoch": 0.24348464456136057,
|
| 1647 |
+
"grad_norm": 0.19187535345554352,
|
| 1648 |
+
"learning_rate": 1.3396495660566651e-05,
|
| 1649 |
+
"loss": 0.1427,
|
| 1650 |
+
"step": 1990
|
| 1651 |
+
},
|
| 1652 |
+
{
|
| 1653 |
+
"epoch": 0.2447081854888046,
|
| 1654 |
+
"grad_norm": 0.10940376669168472,
|
| 1655 |
+
"learning_rate": 1.3319464250424068e-05,
|
| 1656 |
+
"loss": 0.1439,
|
| 1657 |
+
"step": 2000
|
| 1658 |
+
},
|
| 1659 |
+
{
|
| 1660 |
+
"epoch": 0.2447081854888046,
|
| 1661 |
+
"eval_accuracy": 0.97729,
|
| 1662 |
+
"eval_f1_weighted": 0.9772907740157332,
|
| 1663 |
+
"eval_loss": 0.01969166286289692,
|
| 1664 |
+
"eval_macro_f1": 0.9772431978486675,
|
| 1665 |
+
"eval_precision": 0.9772926495694174,
|
| 1666 |
+
"eval_recall": 0.97729,
|
| 1667 |
+
"eval_runtime": 660.9031,
|
| 1668 |
+
"eval_samples_per_second": 151.308,
|
| 1669 |
+
"eval_steps_per_second": 1.183,
|
| 1670 |
+
"step": 2000
|
| 1671 |
+
},
|
| 1672 |
+
{
|
| 1673 |
+
"epoch": 0.24593172641624864,
|
| 1674 |
+
"grad_norm": 0.07293081283569336,
|
| 1675 |
+
"learning_rate": 1.3242210834826575e-05,
|
| 1676 |
+
"loss": 0.1322,
|
| 1677 |
+
"step": 2010
|
| 1678 |
+
},
|
| 1679 |
+
{
|
| 1680 |
+
"epoch": 0.24715526734369264,
|
| 1681 |
+
"grad_norm": 0.1582045555114746,
|
| 1682 |
+
"learning_rate": 1.31647405804753e-05,
|
| 1683 |
+
"loss": 0.1273,
|
| 1684 |
+
"step": 2020
|
| 1685 |
+
},
|
| 1686 |
+
{
|
| 1687 |
+
"epoch": 0.24837880827113668,
|
| 1688 |
+
"grad_norm": 0.29221460223197937,
|
| 1689 |
+
"learning_rate": 1.3087058668573521e-05,
|
| 1690 |
+
"loss": 0.1303,
|
| 1691 |
+
"step": 2030
|
| 1692 |
+
},
|
| 1693 |
+
{
|
| 1694 |
+
"epoch": 0.24960234919858068,
|
| 1695 |
+
"grad_norm": 0.17140211164951324,
|
| 1696 |
+
"learning_rate": 1.3009170294480149e-05,
|
| 1697 |
+
"loss": 0.1394,
|
| 1698 |
+
"step": 2040
|
| 1699 |
+
},
|
| 1700 |
+
{
|
| 1701 |
+
"epoch": 0.25082589012602474,
|
| 1702 |
+
"grad_norm": 0.3539097011089325,
|
| 1703 |
+
"learning_rate": 1.2931080667362276e-05,
|
| 1704 |
+
"loss": 0.1331,
|
| 1705 |
+
"step": 2050
|
| 1706 |
+
},
|
| 1707 |
+
{
|
| 1708 |
+
"epoch": 0.2520494310534687,
|
| 1709 |
+
"grad_norm": 0.23371531069278717,
|
| 1710 |
+
"learning_rate": 1.2852795009846775e-05,
|
| 1711 |
+
"loss": 0.1238,
|
| 1712 |
+
"step": 2060
|
| 1713 |
+
},
|
| 1714 |
+
{
|
| 1715 |
+
"epoch": 0.25327297198091275,
|
| 1716 |
+
"grad_norm": 0.21888205409049988,
|
| 1717 |
+
"learning_rate": 1.2774318557671009e-05,
|
| 1718 |
+
"loss": 0.1509,
|
| 1719 |
+
"step": 2070
|
| 1720 |
+
},
|
| 1721 |
+
{
|
| 1722 |
+
"epoch": 0.2544965129083568,
|
| 1723 |
+
"grad_norm": 0.4899176359176636,
|
| 1724 |
+
"learning_rate": 1.2695656559332671e-05,
|
| 1725 |
+
"loss": 0.1461,
|
| 1726 |
+
"step": 2080
|
| 1727 |
+
},
|
| 1728 |
+
{
|
| 1729 |
+
"epoch": 0.2557200538358008,
|
| 1730 |
+
"grad_norm": 0.1515502780675888,
|
| 1731 |
+
"learning_rate": 1.261681427573877e-05,
|
| 1732 |
+
"loss": 0.1331,
|
| 1733 |
+
"step": 2090
|
| 1734 |
+
},
|
| 1735 |
+
{
|
| 1736 |
+
"epoch": 0.25694359476324485,
|
| 1737 |
+
"grad_norm": 0.13595114648342133,
|
| 1738 |
+
"learning_rate": 1.2537796979853764e-05,
|
| 1739 |
+
"loss": 0.1484,
|
| 1740 |
+
"step": 2100
|
| 1741 |
+
},
|
| 1742 |
+
{
|
| 1743 |
+
"epoch": 0.25694359476324485,
|
| 1744 |
+
"eval_accuracy": 0.97769,
|
| 1745 |
+
"eval_f1_weighted": 0.9776954901619872,
|
| 1746 |
+
"eval_loss": 0.019225435331463814,
|
| 1747 |
+
"eval_macro_f1": 0.9776543362648145,
|
| 1748 |
+
"eval_precision": 0.9777750172865488,
|
| 1749 |
+
"eval_recall": 0.97769,
|
| 1750 |
+
"eval_runtime": 660.729,
|
| 1751 |
+
"eval_samples_per_second": 151.348,
|
| 1752 |
+
"eval_steps_per_second": 1.184,
|
| 1753 |
+
"step": 2100
|
| 1754 |
+
},
|
| 1755 |
+
{
|
| 1756 |
+
"epoch": 0.25816713569068883,
|
| 1757 |
+
"grad_norm": 0.14403671026229858,
|
| 1758 |
+
"learning_rate": 1.2458609956346927e-05,
|
| 1759 |
+
"loss": 0.1386,
|
| 1760 |
+
"step": 2110
|
| 1761 |
+
},
|
| 1762 |
+
{
|
| 1763 |
+
"epoch": 0.25939067661813286,
|
| 1764 |
+
"grad_norm": 0.28055140376091003,
|
| 1765 |
+
"learning_rate": 1.23792585012389e-05,
|
| 1766 |
+
"loss": 0.1331,
|
| 1767 |
+
"step": 2120
|
| 1768 |
+
},
|
| 1769 |
+
{
|
| 1770 |
+
"epoch": 0.2606142175455769,
|
| 1771 |
+
"grad_norm": 0.15281730890274048,
|
| 1772 |
+
"learning_rate": 1.2299747921547485e-05,
|
| 1773 |
+
"loss": 0.1428,
|
| 1774 |
+
"step": 2130
|
| 1775 |
+
},
|
| 1776 |
+
{
|
| 1777 |
+
"epoch": 0.26183775847302093,
|
| 1778 |
+
"grad_norm": 0.2775639593601227,
|
| 1779 |
+
"learning_rate": 1.2220083534932731e-05,
|
| 1780 |
+
"loss": 0.1292,
|
| 1781 |
+
"step": 2140
|
| 1782 |
+
},
|
| 1783 |
+
{
|
| 1784 |
+
"epoch": 0.26306129940046497,
|
| 1785 |
+
"grad_norm": 0.23231762647628784,
|
| 1786 |
+
"learning_rate": 1.2140270669341272e-05,
|
| 1787 |
+
"loss": 0.1278,
|
| 1788 |
+
"step": 2150
|
| 1789 |
+
},
|
| 1790 |
+
{
|
| 1791 |
+
"epoch": 0.26428484032790894,
|
| 1792 |
+
"grad_norm": 0.15349626541137695,
|
| 1793 |
+
"learning_rate": 1.2060314662650008e-05,
|
| 1794 |
+
"loss": 0.1289,
|
| 1795 |
+
"step": 2160
|
| 1796 |
+
},
|
| 1797 |
+
{
|
| 1798 |
+
"epoch": 0.265508381255353,
|
| 1799 |
+
"grad_norm": 0.270783394575119,
|
| 1800 |
+
"learning_rate": 1.1980220862309097e-05,
|
| 1801 |
+
"loss": 0.1353,
|
| 1802 |
+
"step": 2170
|
| 1803 |
+
},
|
| 1804 |
+
{
|
| 1805 |
+
"epoch": 0.266731922182797,
|
| 1806 |
+
"grad_norm": 0.25189876556396484,
|
| 1807 |
+
"learning_rate": 1.1899994624984324e-05,
|
| 1808 |
+
"loss": 0.128,
|
| 1809 |
+
"step": 2180
|
| 1810 |
+
},
|
| 1811 |
+
{
|
| 1812 |
+
"epoch": 0.26795546311024104,
|
| 1813 |
+
"grad_norm": 0.3517930805683136,
|
| 1814 |
+
"learning_rate": 1.1819641316198856e-05,
|
| 1815 |
+
"loss": 0.1368,
|
| 1816 |
+
"step": 2190
|
| 1817 |
+
},
|
| 1818 |
+
{
|
| 1819 |
+
"epoch": 0.2691790040376851,
|
| 1820 |
+
"grad_norm": 0.1853000968694687,
|
| 1821 |
+
"learning_rate": 1.173916630997437e-05,
|
| 1822 |
+
"loss": 0.1261,
|
| 1823 |
+
"step": 2200
|
| 1824 |
+
},
|
| 1825 |
+
{
|
| 1826 |
+
"epoch": 0.2691790040376851,
|
| 1827 |
+
"eval_accuracy": 0.96989,
|
| 1828 |
+
"eval_f1_weighted": 0.9699058894297731,
|
| 1829 |
+
"eval_loss": 0.027551453560590744,
|
| 1830 |
+
"eval_macro_f1": 0.9698711242322156,
|
| 1831 |
+
"eval_precision": 0.97078292839073,
|
| 1832 |
+
"eval_recall": 0.96989,
|
| 1833 |
+
"eval_runtime": 660.4203,
|
| 1834 |
+
"eval_samples_per_second": 151.419,
|
| 1835 |
+
"eval_steps_per_second": 1.184,
|
| 1836 |
+
"step": 2200
|
| 1837 |
+
}
|
| 1838 |
+
],
|
| 1839 |
+
"logging_steps": 10,
|
| 1840 |
+
"max_steps": 8173,
|
| 1841 |
+
"num_input_tokens_seen": 0,
|
| 1842 |
+
"num_train_epochs": 1,
|
| 1843 |
+
"save_steps": 100,
|
| 1844 |
+
"stateful_callbacks": {
|
| 1845 |
+
"EarlyStoppingCallback": {
|
| 1846 |
+
"args": {
|
| 1847 |
+
"early_stopping_patience": 3,
|
| 1848 |
+
"early_stopping_threshold": 0.0
|
| 1849 |
+
},
|
| 1850 |
+
"attributes": {
|
| 1851 |
+
"early_stopping_patience_counter": 1
|
| 1852 |
+
}
|
| 1853 |
+
},
|
| 1854 |
+
"TrainerControl": {
|
| 1855 |
+
"args": {
|
| 1856 |
+
"should_epoch_stop": false,
|
| 1857 |
+
"should_evaluate": false,
|
| 1858 |
+
"should_log": false,
|
| 1859 |
+
"should_save": true,
|
| 1860 |
+
"should_training_stop": false
|
| 1861 |
+
},
|
| 1862 |
+
"attributes": {}
|
| 1863 |
+
}
|
| 1864 |
+
},
|
| 1865 |
+
"total_flos": 3.734350177304909e+16,
|
| 1866 |
+
"train_batch_size": 64,
|
| 1867 |
+
"trial_name": null,
|
| 1868 |
+
"trial_params": null
|
| 1869 |
+
}
|
checkpoint-2200/training_args.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:34af5433a3002793dc6233ee444a7f3918508304d82b9638eca4b9f53208c9bf
|
| 3 |
+
size 5841
|
checkpoint-2200/vocab.json
ADDED
|
The diff for this file is too large to render.
See raw diff
|
|
|