commit files to HF hub
Browse files- added_tokens.json +1 -0
- config.json +43 -0
- merges.txt +0 -0
- optimizer.pt +3 -0
- pytorch_model.bin +3 -0
- rng_state.pth +3 -0
- scheduler.pt +3 -0
- special_tokens_map.json +1 -0
- tokenizer.json +0 -0
- tokenizer_config.json +1 -0
- trainer_state.json +970 -0
- training_args.bin +3 -0
- vocab.json +0 -0
added_tokens.json
ADDED
|
@@ -0,0 +1 @@
|
|
|
|
|
|
|
| 1 |
+
{"[service-Negative]": 50277, "[food-Negative]": 50265, "[ambiance-Positive]": 50270, "[service-no majority]": 50280, "[noise-unknown]": 50275, "[noise-Positive]": 50274, "[service-Positive]": 50278, "[service-unknown]": 50279, "[food-unknown]": 50267, "[noise-no majority]": 50276, "[food-no majority]": 50268, "[ambiance-Negative]": 50269, "[ambiance-no majority]": 50272, "[food-Positive]": 50266, "[noise-Negative]": 50273, "[ambiance-unknown]": 50271}
|
config.json
ADDED
|
@@ -0,0 +1,43 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"_name_or_path": "CEBaB/roberta-base.CEBaB.sa.5-class.exclusive.seed_42",
|
| 3 |
+
"architectures": [
|
| 4 |
+
"RobertaForFactualCounterfactualSequenceClassification"
|
| 5 |
+
],
|
| 6 |
+
"attention_probs_dropout_prob": 0.1,
|
| 7 |
+
"bos_token_id": 0,
|
| 8 |
+
"classifier_dropout": null,
|
| 9 |
+
"eos_token_id": 2,
|
| 10 |
+
"finetuning_task": "opentable",
|
| 11 |
+
"hidden_act": "gelu",
|
| 12 |
+
"hidden_dropout_prob": 0.1,
|
| 13 |
+
"hidden_size": 768,
|
| 14 |
+
"id2label": {
|
| 15 |
+
"0": 0,
|
| 16 |
+
"1": 1,
|
| 17 |
+
"2": 2,
|
| 18 |
+
"3": 3,
|
| 19 |
+
"4": 4
|
| 20 |
+
},
|
| 21 |
+
"initializer_range": 0.02,
|
| 22 |
+
"intermediate_size": 3072,
|
| 23 |
+
"label2id": {
|
| 24 |
+
"0": 0,
|
| 25 |
+
"1": 1,
|
| 26 |
+
"2": 2,
|
| 27 |
+
"3": 3,
|
| 28 |
+
"4": 4
|
| 29 |
+
},
|
| 30 |
+
"layer_norm_eps": 1e-05,
|
| 31 |
+
"max_position_embeddings": 514,
|
| 32 |
+
"model_type": "roberta",
|
| 33 |
+
"num_attention_heads": 12,
|
| 34 |
+
"num_hidden_layers": 12,
|
| 35 |
+
"pad_token_id": 1,
|
| 36 |
+
"position_embedding_type": "absolute",
|
| 37 |
+
"problem_type": "single_label_classification",
|
| 38 |
+
"torch_dtype": "float32",
|
| 39 |
+
"transformers_version": "4.18.0",
|
| 40 |
+
"type_vocab_size": 1,
|
| 41 |
+
"use_cache": true,
|
| 42 |
+
"vocab_size": 50281
|
| 43 |
+
}
|
merges.txt
ADDED
|
The diff for this file is too large to render.
See raw diff
|
|
|
optimizer.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:b1272b73b7150bd91b6a3fc54bc8c88057b396e499df6e8b98b301d18ff97f01
|
| 3 |
+
size 997410589
|
pytorch_model.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:9de110fdf495325181cdf46fc170c92149c5fd7ebb7cfff776f7446bd8e114a5
|
| 3 |
+
size 498718701
|
rng_state.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:120e66cfa1c09184895052ce9bbe12ac6610b030bdf1140d8cc7cfde7ac65171
|
| 3 |
+
size 15523
|
scheduler.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:7e82eb7aef9bd2c8c91fe92b0cb56cc824f3f229a56996ccdacc517e15ec8128
|
| 3 |
+
size 623
|
special_tokens_map.json
ADDED
|
@@ -0,0 +1 @@
|
|
|
|
|
|
|
| 1 |
+
{"bos_token": "<s>", "eos_token": "</s>", "unk_token": "<unk>", "sep_token": "</s>", "pad_token": "<pad>", "cls_token": "<s>", "mask_token": {"content": "<mask>", "single_word": false, "lstrip": true, "rstrip": false, "normalized": false}, "additional_special_tokens": ["[food-Negative]", "[food-Positive]", "[food-unknown]", "[food-no majority]", "[ambiance-Negative]", "[ambiance-Positive]", "[ambiance-unknown]", "[ambiance-no majority]", "[noise-Negative]", "[noise-Positive]", "[noise-unknown]", "[noise-no majority]", "[service-Negative]", "[service-Positive]", "[service-unknown]", "[service-no majority]"]}
|
tokenizer.json
ADDED
|
The diff for this file is too large to render.
See raw diff
|
|
|
tokenizer_config.json
ADDED
|
@@ -0,0 +1 @@
|
|
|
|
|
|
|
| 1 |
+
{"errors": "replace", "bos_token": "<s>", "eos_token": "</s>", "sep_token": "</s>", "cls_token": "<s>", "unk_token": "<unk>", "pad_token": "<pad>", "mask_token": "<mask>", "add_prefix_space": false, "trim_offsets": true, "model_max_length": 512, "special_tokens_map_file": null, "name_or_path": "roberta-base", "tokenizer_class": "RobertaTokenizer"}
|
trainer_state.json
ADDED
|
@@ -0,0 +1,970 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": 0.44814541935920715,
|
| 3 |
+
"best_model_checkpoint": "model_output/e2e_opentable_5_way__approximate__0-shot__seed-42__roberta-base/checkpoint-1450",
|
| 4 |
+
"epoch": 2.4871355060034306,
|
| 5 |
+
"global_step": 1450,
|
| 6 |
+
"is_hyper_param_search": false,
|
| 7 |
+
"is_local_process_zero": true,
|
| 8 |
+
"is_world_process_zero": true,
|
| 9 |
+
"log_history": [
|
| 10 |
+
{
|
| 11 |
+
"epoch": 0.03,
|
| 12 |
+
"learning_rate": 4.9783362218370885e-05,
|
| 13 |
+
"loss": 15.3848,
|
| 14 |
+
"step": 20
|
| 15 |
+
},
|
| 16 |
+
{
|
| 17 |
+
"epoch": 0.07,
|
| 18 |
+
"learning_rate": 4.956672443674177e-05,
|
| 19 |
+
"loss": 12.8162,
|
| 20 |
+
"step": 40
|
| 21 |
+
},
|
| 22 |
+
{
|
| 23 |
+
"epoch": 0.09,
|
| 24 |
+
"eval_distillation_accuracy_counterfactual": 0.4474089276552078,
|
| 25 |
+
"eval_distillation_accuracy_factual": 0.7721908671113391,
|
| 26 |
+
"eval_distillation_f1_counterfactual": 0.34384085542828363,
|
| 27 |
+
"eval_distillation_f1_factual": 0.7449065740174884,
|
| 28 |
+
"eval_groundtruth_accuracy_counterfactual": 0.43560800410466904,
|
| 29 |
+
"eval_groundtruth_f1_counterfactual": 0.3322658600412939,
|
| 30 |
+
"eval_groundtruth_f1_factual": 0.6497934851385894,
|
| 31 |
+
"eval_icace_cosine": 0.4707597494125366,
|
| 32 |
+
"eval_icace_l2": 0.7290830612182617,
|
| 33 |
+
"eval_icace_normdiff": 0.5084162950515747,
|
| 34 |
+
"eval_loss": 9.263871192932129,
|
| 35 |
+
"eval_runtime": 12.7695,
|
| 36 |
+
"eval_samples_per_second": 305.258,
|
| 37 |
+
"eval_steps_per_second": 2.428,
|
| 38 |
+
"step": 50
|
| 39 |
+
},
|
| 40 |
+
{
|
| 41 |
+
"epoch": 0.1,
|
| 42 |
+
"learning_rate": 4.935008665511265e-05,
|
| 43 |
+
"loss": 11.524,
|
| 44 |
+
"step": 60
|
| 45 |
+
},
|
| 46 |
+
{
|
| 47 |
+
"epoch": 0.14,
|
| 48 |
+
"learning_rate": 4.913344887348354e-05,
|
| 49 |
+
"loss": 10.7187,
|
| 50 |
+
"step": 80
|
| 51 |
+
},
|
| 52 |
+
{
|
| 53 |
+
"epoch": 0.17,
|
| 54 |
+
"learning_rate": 4.891681109185442e-05,
|
| 55 |
+
"loss": 10.3266,
|
| 56 |
+
"step": 100
|
| 57 |
+
},
|
| 58 |
+
{
|
| 59 |
+
"epoch": 0.17,
|
| 60 |
+
"eval_distillation_accuracy_counterfactual": 0.47485890200102615,
|
| 61 |
+
"eval_distillation_accuracy_factual": 0.8653155464340687,
|
| 62 |
+
"eval_distillation_f1_counterfactual": 0.352733121767057,
|
| 63 |
+
"eval_distillation_f1_factual": 0.8531998683010464,
|
| 64 |
+
"eval_groundtruth_accuracy_counterfactual": 0.47691123653155465,
|
| 65 |
+
"eval_groundtruth_f1_counterfactual": 0.3544569665529119,
|
| 66 |
+
"eval_groundtruth_f1_factual": 0.7055142150493701,
|
| 67 |
+
"eval_icace_cosine": 0.46635791659355164,
|
| 68 |
+
"eval_icace_l2": 0.7056604623794556,
|
| 69 |
+
"eval_icace_normdiff": 0.4767073690891266,
|
| 70 |
+
"eval_loss": 8.300187110900879,
|
| 71 |
+
"eval_runtime": 13.1062,
|
| 72 |
+
"eval_samples_per_second": 297.416,
|
| 73 |
+
"eval_steps_per_second": 2.365,
|
| 74 |
+
"step": 100
|
| 75 |
+
},
|
| 76 |
+
{
|
| 77 |
+
"epoch": 0.21,
|
| 78 |
+
"learning_rate": 4.8700173310225307e-05,
|
| 79 |
+
"loss": 9.8902,
|
| 80 |
+
"step": 120
|
| 81 |
+
},
|
| 82 |
+
{
|
| 83 |
+
"epoch": 0.24,
|
| 84 |
+
"learning_rate": 4.848353552859619e-05,
|
| 85 |
+
"loss": 9.9859,
|
| 86 |
+
"step": 140
|
| 87 |
+
},
|
| 88 |
+
{
|
| 89 |
+
"epoch": 0.26,
|
| 90 |
+
"eval_distillation_accuracy_counterfactual": 0.46587993842996406,
|
| 91 |
+
"eval_distillation_accuracy_factual": 0.8286300667008722,
|
| 92 |
+
"eval_distillation_f1_counterfactual": 0.3884433215928239,
|
| 93 |
+
"eval_distillation_f1_factual": 0.8139544299165415,
|
| 94 |
+
"eval_groundtruth_accuracy_counterfactual": 0.46151872755259105,
|
| 95 |
+
"eval_groundtruth_f1_counterfactual": 0.3891775674468404,
|
| 96 |
+
"eval_groundtruth_f1_factual": 0.6795376489414218,
|
| 97 |
+
"eval_icace_cosine": 0.4685736894607544,
|
| 98 |
+
"eval_icace_l2": 0.6997411251068115,
|
| 99 |
+
"eval_icace_normdiff": 0.4409981966018677,
|
| 100 |
+
"eval_loss": 8.554189682006836,
|
| 101 |
+
"eval_runtime": 12.8948,
|
| 102 |
+
"eval_samples_per_second": 302.293,
|
| 103 |
+
"eval_steps_per_second": 2.404,
|
| 104 |
+
"step": 150
|
| 105 |
+
},
|
| 106 |
+
{
|
| 107 |
+
"epoch": 0.27,
|
| 108 |
+
"learning_rate": 4.826689774696707e-05,
|
| 109 |
+
"loss": 9.9401,
|
| 110 |
+
"step": 160
|
| 111 |
+
},
|
| 112 |
+
{
|
| 113 |
+
"epoch": 0.31,
|
| 114 |
+
"learning_rate": 4.8050259965337955e-05,
|
| 115 |
+
"loss": 9.8133,
|
| 116 |
+
"step": 180
|
| 117 |
+
},
|
| 118 |
+
{
|
| 119 |
+
"epoch": 0.34,
|
| 120 |
+
"learning_rate": 4.7833622183708845e-05,
|
| 121 |
+
"loss": 9.5608,
|
| 122 |
+
"step": 200
|
| 123 |
+
},
|
| 124 |
+
{
|
| 125 |
+
"epoch": 0.34,
|
| 126 |
+
"eval_distillation_accuracy_counterfactual": 0.49589533093894306,
|
| 127 |
+
"eval_distillation_accuracy_factual": 0.8476141611082606,
|
| 128 |
+
"eval_distillation_f1_counterfactual": 0.41223146480687944,
|
| 129 |
+
"eval_distillation_f1_factual": 0.839115183452182,
|
| 130 |
+
"eval_groundtruth_accuracy_counterfactual": 0.4979476654694715,
|
| 131 |
+
"eval_groundtruth_f1_counterfactual": 0.4149936071266812,
|
| 132 |
+
"eval_groundtruth_f1_factual": 0.6763507473779132,
|
| 133 |
+
"eval_icace_cosine": 0.461904913187027,
|
| 134 |
+
"eval_icace_l2": 0.6693803668022156,
|
| 135 |
+
"eval_icace_normdiff": 0.4146590232849121,
|
| 136 |
+
"eval_loss": 8.107243537902832,
|
| 137 |
+
"eval_runtime": 13.0158,
|
| 138 |
+
"eval_samples_per_second": 299.481,
|
| 139 |
+
"eval_steps_per_second": 2.382,
|
| 140 |
+
"step": 200
|
| 141 |
+
},
|
| 142 |
+
{
|
| 143 |
+
"epoch": 0.38,
|
| 144 |
+
"learning_rate": 4.761698440207972e-05,
|
| 145 |
+
"loss": 9.6041,
|
| 146 |
+
"step": 220
|
| 147 |
+
},
|
| 148 |
+
{
|
| 149 |
+
"epoch": 0.41,
|
| 150 |
+
"learning_rate": 4.740034662045061e-05,
|
| 151 |
+
"loss": 9.4542,
|
| 152 |
+
"step": 240
|
| 153 |
+
},
|
| 154 |
+
{
|
| 155 |
+
"epoch": 0.43,
|
| 156 |
+
"eval_distillation_accuracy_counterfactual": 0.5123140071831709,
|
| 157 |
+
"eval_distillation_accuracy_factual": 0.8365828630066701,
|
| 158 |
+
"eval_distillation_f1_counterfactual": 0.41742508528245137,
|
| 159 |
+
"eval_distillation_f1_factual": 0.8170598509732152,
|
| 160 |
+
"eval_groundtruth_accuracy_counterfactual": 0.4923037455105182,
|
| 161 |
+
"eval_groundtruth_f1_counterfactual": 0.39565449482548276,
|
| 162 |
+
"eval_groundtruth_f1_factual": 0.695773274631551,
|
| 163 |
+
"eval_icace_cosine": 0.4568592607975006,
|
| 164 |
+
"eval_icace_l2": 0.6699839234352112,
|
| 165 |
+
"eval_icace_normdiff": 0.4179259240627289,
|
| 166 |
+
"eval_loss": 8.091322898864746,
|
| 167 |
+
"eval_runtime": 13.0177,
|
| 168 |
+
"eval_samples_per_second": 299.438,
|
| 169 |
+
"eval_steps_per_second": 2.381,
|
| 170 |
+
"step": 250
|
| 171 |
+
},
|
| 172 |
+
{
|
| 173 |
+
"epoch": 0.45,
|
| 174 |
+
"learning_rate": 4.7183708838821494e-05,
|
| 175 |
+
"loss": 9.4606,
|
| 176 |
+
"step": 260
|
| 177 |
+
},
|
| 178 |
+
{
|
| 179 |
+
"epoch": 0.48,
|
| 180 |
+
"learning_rate": 4.6967071057192376e-05,
|
| 181 |
+
"loss": 9.4165,
|
| 182 |
+
"step": 280
|
| 183 |
+
},
|
| 184 |
+
{
|
| 185 |
+
"epoch": 0.51,
|
| 186 |
+
"learning_rate": 4.675043327556326e-05,
|
| 187 |
+
"loss": 9.3996,
|
| 188 |
+
"step": 300
|
| 189 |
+
},
|
| 190 |
+
{
|
| 191 |
+
"epoch": 0.51,
|
| 192 |
+
"eval_distillation_accuracy_counterfactual": 0.5161621344279117,
|
| 193 |
+
"eval_distillation_accuracy_factual": 0.8445356593124679,
|
| 194 |
+
"eval_distillation_f1_counterfactual": 0.43140398163976057,
|
| 195 |
+
"eval_distillation_f1_factual": 0.8243538167177713,
|
| 196 |
+
"eval_groundtruth_accuracy_counterfactual": 0.5071831708568497,
|
| 197 |
+
"eval_groundtruth_f1_counterfactual": 0.4211696295155843,
|
| 198 |
+
"eval_groundtruth_f1_factual": 0.6804350127556201,
|
| 199 |
+
"eval_icace_cosine": 0.4562739133834839,
|
| 200 |
+
"eval_icace_l2": 0.671053409576416,
|
| 201 |
+
"eval_icace_normdiff": 0.42435094714164734,
|
| 202 |
+
"eval_loss": 7.912776470184326,
|
| 203 |
+
"eval_runtime": 13.0458,
|
| 204 |
+
"eval_samples_per_second": 298.794,
|
| 205 |
+
"eval_steps_per_second": 2.376,
|
| 206 |
+
"step": 300
|
| 207 |
+
},
|
| 208 |
+
{
|
| 209 |
+
"epoch": 0.55,
|
| 210 |
+
"learning_rate": 4.653379549393415e-05,
|
| 211 |
+
"loss": 9.2297,
|
| 212 |
+
"step": 320
|
| 213 |
+
},
|
| 214 |
+
{
|
| 215 |
+
"epoch": 0.58,
|
| 216 |
+
"learning_rate": 4.6317157712305025e-05,
|
| 217 |
+
"loss": 9.2203,
|
| 218 |
+
"step": 340
|
| 219 |
+
},
|
| 220 |
+
{
|
| 221 |
+
"epoch": 0.6,
|
| 222 |
+
"eval_distillation_accuracy_counterfactual": 0.5179579271421242,
|
| 223 |
+
"eval_distillation_accuracy_factual": 0.8155464340687532,
|
| 224 |
+
"eval_distillation_f1_counterfactual": 0.4682566845228989,
|
| 225 |
+
"eval_distillation_f1_factual": 0.8006032026406558,
|
| 226 |
+
"eval_groundtruth_accuracy_counterfactual": 0.5020523345305284,
|
| 227 |
+
"eval_groundtruth_f1_counterfactual": 0.45492108472946924,
|
| 228 |
+
"eval_groundtruth_f1_factual": 0.666677263616981,
|
| 229 |
+
"eval_icace_cosine": 0.455741286277771,
|
| 230 |
+
"eval_icace_l2": 0.6516066789627075,
|
| 231 |
+
"eval_icace_normdiff": 0.4051017761230469,
|
| 232 |
+
"eval_loss": 8.105178833007812,
|
| 233 |
+
"eval_runtime": 12.9117,
|
| 234 |
+
"eval_samples_per_second": 301.897,
|
| 235 |
+
"eval_steps_per_second": 2.401,
|
| 236 |
+
"step": 350
|
| 237 |
+
},
|
| 238 |
+
{
|
| 239 |
+
"epoch": 0.62,
|
| 240 |
+
"learning_rate": 4.6100519930675915e-05,
|
| 241 |
+
"loss": 9.0931,
|
| 242 |
+
"step": 360
|
| 243 |
+
},
|
| 244 |
+
{
|
| 245 |
+
"epoch": 0.65,
|
| 246 |
+
"learning_rate": 4.58838821490468e-05,
|
| 247 |
+
"loss": 8.9297,
|
| 248 |
+
"step": 380
|
| 249 |
+
},
|
| 250 |
+
{
|
| 251 |
+
"epoch": 0.69,
|
| 252 |
+
"learning_rate": 4.566724436741768e-05,
|
| 253 |
+
"loss": 8.8054,
|
| 254 |
+
"step": 400
|
| 255 |
+
},
|
| 256 |
+
{
|
| 257 |
+
"epoch": 0.69,
|
| 258 |
+
"eval_distillation_accuracy_counterfactual": 0.5053873781426372,
|
| 259 |
+
"eval_distillation_accuracy_factual": 0.8222165212929707,
|
| 260 |
+
"eval_distillation_f1_counterfactual": 0.38822036619209827,
|
| 261 |
+
"eval_distillation_f1_factual": 0.8117129824518028,
|
| 262 |
+
"eval_groundtruth_accuracy_counterfactual": 0.5010261672652643,
|
| 263 |
+
"eval_groundtruth_f1_counterfactual": 0.38820947576310105,
|
| 264 |
+
"eval_groundtruth_f1_factual": 0.6914330659012916,
|
| 265 |
+
"eval_icace_cosine": 0.45915645360946655,
|
| 266 |
+
"eval_icace_l2": 0.6688275933265686,
|
| 267 |
+
"eval_icace_normdiff": 0.4207516312599182,
|
| 268 |
+
"eval_loss": 8.215511322021484,
|
| 269 |
+
"eval_runtime": 13.0382,
|
| 270 |
+
"eval_samples_per_second": 298.967,
|
| 271 |
+
"eval_steps_per_second": 2.378,
|
| 272 |
+
"step": 400
|
| 273 |
+
},
|
| 274 |
+
{
|
| 275 |
+
"epoch": 0.72,
|
| 276 |
+
"learning_rate": 4.5450606585788563e-05,
|
| 277 |
+
"loss": 9.0175,
|
| 278 |
+
"step": 420
|
| 279 |
+
},
|
| 280 |
+
{
|
| 281 |
+
"epoch": 0.75,
|
| 282 |
+
"learning_rate": 4.5233968804159446e-05,
|
| 283 |
+
"loss": 9.0479,
|
| 284 |
+
"step": 440
|
| 285 |
+
},
|
| 286 |
+
{
|
| 287 |
+
"epoch": 0.77,
|
| 288 |
+
"eval_distillation_accuracy_counterfactual": 0.5161621344279117,
|
| 289 |
+
"eval_distillation_accuracy_factual": 0.8322216521292971,
|
| 290 |
+
"eval_distillation_f1_counterfactual": 0.43163093824456045,
|
| 291 |
+
"eval_distillation_f1_factual": 0.8179434971552079,
|
| 292 |
+
"eval_groundtruth_accuracy_counterfactual": 0.5164186762442278,
|
| 293 |
+
"eval_groundtruth_f1_counterfactual": 0.43368043593910377,
|
| 294 |
+
"eval_groundtruth_f1_factual": 0.7010946530829851,
|
| 295 |
+
"eval_icace_cosine": 0.46967530250549316,
|
| 296 |
+
"eval_icace_l2": 0.6567780375480652,
|
| 297 |
+
"eval_icace_normdiff": 0.38229528069496155,
|
| 298 |
+
"eval_loss": 8.273370742797852,
|
| 299 |
+
"eval_runtime": 12.9013,
|
| 300 |
+
"eval_samples_per_second": 302.139,
|
| 301 |
+
"eval_steps_per_second": 2.403,
|
| 302 |
+
"step": 450
|
| 303 |
+
},
|
| 304 |
+
{
|
| 305 |
+
"epoch": 0.79,
|
| 306 |
+
"learning_rate": 4.501733102253033e-05,
|
| 307 |
+
"loss": 8.6738,
|
| 308 |
+
"step": 460
|
| 309 |
+
},
|
| 310 |
+
{
|
| 311 |
+
"epoch": 0.82,
|
| 312 |
+
"learning_rate": 4.480069324090121e-05,
|
| 313 |
+
"loss": 8.8553,
|
| 314 |
+
"step": 480
|
| 315 |
+
},
|
| 316 |
+
{
|
| 317 |
+
"epoch": 0.86,
|
| 318 |
+
"learning_rate": 4.45840554592721e-05,
|
| 319 |
+
"loss": 8.7079,
|
| 320 |
+
"step": 500
|
| 321 |
+
},
|
| 322 |
+
{
|
| 323 |
+
"epoch": 0.86,
|
| 324 |
+
"eval_distillation_accuracy_counterfactual": 0.5023088763468445,
|
| 325 |
+
"eval_distillation_accuracy_factual": 0.8260646485377117,
|
| 326 |
+
"eval_distillation_f1_counterfactual": 0.39076709834402334,
|
| 327 |
+
"eval_distillation_f1_factual": 0.8131187804629831,
|
| 328 |
+
"eval_groundtruth_accuracy_counterfactual": 0.4974345818368394,
|
| 329 |
+
"eval_groundtruth_f1_counterfactual": 0.38463954865815575,
|
| 330 |
+
"eval_groundtruth_f1_factual": 0.7000450508607393,
|
| 331 |
+
"eval_icace_cosine": 0.474086195230484,
|
| 332 |
+
"eval_icace_l2": 0.6828737258911133,
|
| 333 |
+
"eval_icace_normdiff": 0.40336331725120544,
|
| 334 |
+
"eval_loss": 8.616800308227539,
|
| 335 |
+
"eval_runtime": 13.0775,
|
| 336 |
+
"eval_samples_per_second": 298.07,
|
| 337 |
+
"eval_steps_per_second": 2.37,
|
| 338 |
+
"step": 500
|
| 339 |
+
},
|
| 340 |
+
{
|
| 341 |
+
"epoch": 0.89,
|
| 342 |
+
"learning_rate": 4.436741767764298e-05,
|
| 343 |
+
"loss": 8.795,
|
| 344 |
+
"step": 520
|
| 345 |
+
},
|
| 346 |
+
{
|
| 347 |
+
"epoch": 0.93,
|
| 348 |
+
"learning_rate": 4.415077989601387e-05,
|
| 349 |
+
"loss": 8.8781,
|
| 350 |
+
"step": 540
|
| 351 |
+
},
|
| 352 |
+
{
|
| 353 |
+
"epoch": 0.94,
|
| 354 |
+
"eval_distillation_accuracy_counterfactual": 0.5330938943047717,
|
| 355 |
+
"eval_distillation_accuracy_factual": 0.8201641867624423,
|
| 356 |
+
"eval_distillation_f1_counterfactual": 0.4628830838776178,
|
| 357 |
+
"eval_distillation_f1_factual": 0.8086604463285127,
|
| 358 |
+
"eval_groundtruth_accuracy_counterfactual": 0.5389943560800411,
|
| 359 |
+
"eval_groundtruth_f1_counterfactual": 0.47066366125954884,
|
| 360 |
+
"eval_groundtruth_f1_factual": 0.7039950616687438,
|
| 361 |
+
"eval_icace_cosine": 0.4582160413265228,
|
| 362 |
+
"eval_icace_l2": 0.6311160326004028,
|
| 363 |
+
"eval_icace_normdiff": 0.3835700452327728,
|
| 364 |
+
"eval_loss": 7.966909408569336,
|
| 365 |
+
"eval_runtime": 12.9904,
|
| 366 |
+
"eval_samples_per_second": 300.068,
|
| 367 |
+
"eval_steps_per_second": 2.386,
|
| 368 |
+
"step": 550
|
| 369 |
+
},
|
| 370 |
+
{
|
| 371 |
+
"epoch": 0.96,
|
| 372 |
+
"learning_rate": 4.393414211438475e-05,
|
| 373 |
+
"loss": 8.8926,
|
| 374 |
+
"step": 560
|
| 375 |
+
},
|
| 376 |
+
{
|
| 377 |
+
"epoch": 0.99,
|
| 378 |
+
"learning_rate": 4.371750433275563e-05,
|
| 379 |
+
"loss": 8.704,
|
| 380 |
+
"step": 580
|
| 381 |
+
},
|
| 382 |
+
{
|
| 383 |
+
"epoch": 1.03,
|
| 384 |
+
"learning_rate": 4.3500866551126516e-05,
|
| 385 |
+
"loss": 8.5031,
|
| 386 |
+
"step": 600
|
| 387 |
+
},
|
| 388 |
+
{
|
| 389 |
+
"epoch": 1.03,
|
| 390 |
+
"eval_distillation_accuracy_counterfactual": 0.513083632632119,
|
| 391 |
+
"eval_distillation_accuracy_factual": 0.822986146741919,
|
| 392 |
+
"eval_distillation_f1_counterfactual": 0.44551358939731,
|
| 393 |
+
"eval_distillation_f1_factual": 0.8106574040803418,
|
| 394 |
+
"eval_groundtruth_accuracy_counterfactual": 0.5100051308363264,
|
| 395 |
+
"eval_groundtruth_f1_counterfactual": 0.44647837744933117,
|
| 396 |
+
"eval_groundtruth_f1_factual": 0.7133952359814255,
|
| 397 |
+
"eval_icace_cosine": 0.46173933148384094,
|
| 398 |
+
"eval_icace_l2": 0.6598149538040161,
|
| 399 |
+
"eval_icace_normdiff": 0.4006192982196808,
|
| 400 |
+
"eval_loss": 8.42484188079834,
|
| 401 |
+
"eval_runtime": 12.9066,
|
| 402 |
+
"eval_samples_per_second": 302.016,
|
| 403 |
+
"eval_steps_per_second": 2.402,
|
| 404 |
+
"step": 600
|
| 405 |
+
},
|
| 406 |
+
{
|
| 407 |
+
"epoch": 1.06,
|
| 408 |
+
"learning_rate": 4.3284228769497406e-05,
|
| 409 |
+
"loss": 8.3235,
|
| 410 |
+
"step": 620
|
| 411 |
+
},
|
| 412 |
+
{
|
| 413 |
+
"epoch": 1.1,
|
| 414 |
+
"learning_rate": 4.306759098786828e-05,
|
| 415 |
+
"loss": 8.3831,
|
| 416 |
+
"step": 640
|
| 417 |
+
},
|
| 418 |
+
{
|
| 419 |
+
"epoch": 1.11,
|
| 420 |
+
"eval_distillation_accuracy_counterfactual": 0.5289892252437147,
|
| 421 |
+
"eval_distillation_accuracy_factual": 0.8260646485377117,
|
| 422 |
+
"eval_distillation_f1_counterfactual": 0.47012361221009824,
|
| 423 |
+
"eval_distillation_f1_factual": 0.8064982655163991,
|
| 424 |
+
"eval_groundtruth_accuracy_counterfactual": 0.5302719343252951,
|
| 425 |
+
"eval_groundtruth_f1_counterfactual": 0.47179239129259665,
|
| 426 |
+
"eval_groundtruth_f1_factual": 0.6996350113744447,
|
| 427 |
+
"eval_icace_cosine": 0.4589817523956299,
|
| 428 |
+
"eval_icace_l2": 0.6445746421813965,
|
| 429 |
+
"eval_icace_normdiff": 0.39702358841896057,
|
| 430 |
+
"eval_loss": 8.185053825378418,
|
| 431 |
+
"eval_runtime": 12.9149,
|
| 432 |
+
"eval_samples_per_second": 301.822,
|
| 433 |
+
"eval_steps_per_second": 2.4,
|
| 434 |
+
"step": 650
|
| 435 |
+
},
|
| 436 |
+
{
|
| 437 |
+
"epoch": 1.13,
|
| 438 |
+
"learning_rate": 4.285095320623917e-05,
|
| 439 |
+
"loss": 8.5789,
|
| 440 |
+
"step": 660
|
| 441 |
+
},
|
| 442 |
+
{
|
| 443 |
+
"epoch": 1.17,
|
| 444 |
+
"learning_rate": 4.2634315424610055e-05,
|
| 445 |
+
"loss": 8.4277,
|
| 446 |
+
"step": 680
|
| 447 |
+
},
|
| 448 |
+
{
|
| 449 |
+
"epoch": 1.2,
|
| 450 |
+
"learning_rate": 4.241767764298094e-05,
|
| 451 |
+
"loss": 8.4591,
|
| 452 |
+
"step": 700
|
| 453 |
+
},
|
| 454 |
+
{
|
| 455 |
+
"epoch": 1.2,
|
| 456 |
+
"eval_distillation_accuracy_counterfactual": 0.5071831708568497,
|
| 457 |
+
"eval_distillation_accuracy_factual": 0.8291431503335044,
|
| 458 |
+
"eval_distillation_f1_counterfactual": 0.4229278158258259,
|
| 459 |
+
"eval_distillation_f1_factual": 0.8158920555245694,
|
| 460 |
+
"eval_groundtruth_accuracy_counterfactual": 0.5094920472036942,
|
| 461 |
+
"eval_groundtruth_f1_counterfactual": 0.42468748967332876,
|
| 462 |
+
"eval_groundtruth_f1_factual": 0.7058832006651468,
|
| 463 |
+
"eval_icace_cosine": 0.46126317977905273,
|
| 464 |
+
"eval_icace_l2": 0.6510809659957886,
|
| 465 |
+
"eval_icace_normdiff": 0.39125320315361023,
|
| 466 |
+
"eval_loss": 8.368897438049316,
|
| 467 |
+
"eval_runtime": 12.9121,
|
| 468 |
+
"eval_samples_per_second": 301.886,
|
| 469 |
+
"eval_steps_per_second": 2.401,
|
| 470 |
+
"step": 700
|
| 471 |
+
},
|
| 472 |
+
{
|
| 473 |
+
"epoch": 1.23,
|
| 474 |
+
"learning_rate": 4.220103986135182e-05,
|
| 475 |
+
"loss": 8.4671,
|
| 476 |
+
"step": 720
|
| 477 |
+
},
|
| 478 |
+
{
|
| 479 |
+
"epoch": 1.27,
|
| 480 |
+
"learning_rate": 4.198440207972271e-05,
|
| 481 |
+
"loss": 8.4019,
|
| 482 |
+
"step": 740
|
| 483 |
+
},
|
| 484 |
+
{
|
| 485 |
+
"epoch": 1.29,
|
| 486 |
+
"eval_distillation_accuracy_counterfactual": 0.49281682914315034,
|
| 487 |
+
"eval_distillation_accuracy_factual": 0.827860441251924,
|
| 488 |
+
"eval_distillation_f1_counterfactual": 0.41244951531506935,
|
| 489 |
+
"eval_distillation_f1_factual": 0.8170673706163509,
|
| 490 |
+
"eval_groundtruth_accuracy_counterfactual": 0.4856336582863007,
|
| 491 |
+
"eval_groundtruth_f1_counterfactual": 0.4077764907535967,
|
| 492 |
+
"eval_groundtruth_f1_factual": 0.7051511278749412,
|
| 493 |
+
"eval_icace_cosine": 0.4673001766204834,
|
| 494 |
+
"eval_icace_l2": 0.6646483540534973,
|
| 495 |
+
"eval_icace_normdiff": 0.3972298800945282,
|
| 496 |
+
"eval_loss": 8.459917068481445,
|
| 497 |
+
"eval_runtime": 12.9198,
|
| 498 |
+
"eval_samples_per_second": 301.708,
|
| 499 |
+
"eval_steps_per_second": 2.399,
|
| 500 |
+
"step": 750
|
| 501 |
+
},
|
| 502 |
+
{
|
| 503 |
+
"epoch": 1.3,
|
| 504 |
+
"learning_rate": 4.1767764298093586e-05,
|
| 505 |
+
"loss": 8.5387,
|
| 506 |
+
"step": 760
|
| 507 |
+
},
|
| 508 |
+
{
|
| 509 |
+
"epoch": 1.34,
|
| 510 |
+
"learning_rate": 4.1551126516464476e-05,
|
| 511 |
+
"loss": 8.4688,
|
| 512 |
+
"step": 780
|
| 513 |
+
},
|
| 514 |
+
{
|
| 515 |
+
"epoch": 1.37,
|
| 516 |
+
"learning_rate": 4.133448873483536e-05,
|
| 517 |
+
"loss": 8.5516,
|
| 518 |
+
"step": 800
|
| 519 |
+
},
|
| 520 |
+
{
|
| 521 |
+
"epoch": 1.37,
|
| 522 |
+
"eval_distillation_accuracy_counterfactual": 0.5184710107747563,
|
| 523 |
+
"eval_distillation_accuracy_factual": 0.822986146741919,
|
| 524 |
+
"eval_distillation_f1_counterfactual": 0.4457687437030219,
|
| 525 |
+
"eval_distillation_f1_factual": 0.8022682230012104,
|
| 526 |
+
"eval_groundtruth_accuracy_counterfactual": 0.5159055926115957,
|
| 527 |
+
"eval_groundtruth_f1_counterfactual": 0.4464137100072511,
|
| 528 |
+
"eval_groundtruth_f1_factual": 0.6935766782924615,
|
| 529 |
+
"eval_icace_cosine": 0.45302730798721313,
|
| 530 |
+
"eval_icace_l2": 0.6743680238723755,
|
| 531 |
+
"eval_icace_normdiff": 0.42486244440078735,
|
| 532 |
+
"eval_loss": 8.359912872314453,
|
| 533 |
+
"eval_runtime": 12.9096,
|
| 534 |
+
"eval_samples_per_second": 301.945,
|
| 535 |
+
"eval_steps_per_second": 2.401,
|
| 536 |
+
"step": 800
|
| 537 |
+
},
|
| 538 |
+
{
|
| 539 |
+
"epoch": 1.41,
|
| 540 |
+
"learning_rate": 4.111785095320624e-05,
|
| 541 |
+
"loss": 8.4687,
|
| 542 |
+
"step": 820
|
| 543 |
+
},
|
| 544 |
+
{
|
| 545 |
+
"epoch": 1.44,
|
| 546 |
+
"learning_rate": 4.0901213171577124e-05,
|
| 547 |
+
"loss": 8.3884,
|
| 548 |
+
"step": 840
|
| 549 |
+
},
|
| 550 |
+
{
|
| 551 |
+
"epoch": 1.46,
|
| 552 |
+
"eval_distillation_accuracy_counterfactual": 0.530015392508979,
|
| 553 |
+
"eval_distillation_accuracy_factual": 0.8209338122113905,
|
| 554 |
+
"eval_distillation_f1_counterfactual": 0.44736802933677255,
|
| 555 |
+
"eval_distillation_f1_factual": 0.8048461947798791,
|
| 556 |
+
"eval_groundtruth_accuracy_counterfactual": 0.5323242688558235,
|
| 557 |
+
"eval_groundtruth_f1_counterfactual": 0.45097140064821445,
|
| 558 |
+
"eval_groundtruth_f1_factual": 0.7159403464102823,
|
| 559 |
+
"eval_icace_cosine": 0.4506547152996063,
|
| 560 |
+
"eval_icace_l2": 0.6407148241996765,
|
| 561 |
+
"eval_icace_normdiff": 0.395082026720047,
|
| 562 |
+
"eval_loss": 8.082590103149414,
|
| 563 |
+
"eval_runtime": 12.9483,
|
| 564 |
+
"eval_samples_per_second": 301.044,
|
| 565 |
+
"eval_steps_per_second": 2.394,
|
| 566 |
+
"step": 850
|
| 567 |
+
},
|
| 568 |
+
{
|
| 569 |
+
"epoch": 1.48,
|
| 570 |
+
"learning_rate": 4.068457538994801e-05,
|
| 571 |
+
"loss": 8.4295,
|
| 572 |
+
"step": 860
|
| 573 |
+
},
|
| 574 |
+
{
|
| 575 |
+
"epoch": 1.51,
|
| 576 |
+
"learning_rate": 4.04679376083189e-05,
|
| 577 |
+
"loss": 8.3828,
|
| 578 |
+
"step": 880
|
| 579 |
+
},
|
| 580 |
+
{
|
| 581 |
+
"epoch": 1.54,
|
| 582 |
+
"learning_rate": 4.025129982668977e-05,
|
| 583 |
+
"loss": 8.4004,
|
| 584 |
+
"step": 900
|
| 585 |
+
},
|
| 586 |
+
{
|
| 587 |
+
"epoch": 1.54,
|
| 588 |
+
"eval_distillation_accuracy_counterfactual": 0.5220625962031811,
|
| 589 |
+
"eval_distillation_accuracy_factual": 0.8083632632119035,
|
| 590 |
+
"eval_distillation_f1_counterfactual": 0.44838757222370995,
|
| 591 |
+
"eval_distillation_f1_factual": 0.7919448574347471,
|
| 592 |
+
"eval_groundtruth_accuracy_counterfactual": 0.5243714725500257,
|
| 593 |
+
"eval_groundtruth_f1_counterfactual": 0.44872773811316585,
|
| 594 |
+
"eval_groundtruth_f1_factual": 0.7046094122055937,
|
| 595 |
+
"eval_icace_cosine": 0.4570869207382202,
|
| 596 |
+
"eval_icace_l2": 0.6522126197814941,
|
| 597 |
+
"eval_icace_normdiff": 0.406441867351532,
|
| 598 |
+
"eval_loss": 8.266593933105469,
|
| 599 |
+
"eval_runtime": 12.9504,
|
| 600 |
+
"eval_samples_per_second": 300.995,
|
| 601 |
+
"eval_steps_per_second": 2.394,
|
| 602 |
+
"step": 900
|
| 603 |
+
},
|
| 604 |
+
{
|
| 605 |
+
"epoch": 1.58,
|
| 606 |
+
"learning_rate": 4.003466204506066e-05,
|
| 607 |
+
"loss": 8.335,
|
| 608 |
+
"step": 920
|
| 609 |
+
},
|
| 610 |
+
{
|
| 611 |
+
"epoch": 1.61,
|
| 612 |
+
"learning_rate": 3.9818024263431546e-05,
|
| 613 |
+
"loss": 8.3741,
|
| 614 |
+
"step": 940
|
| 615 |
+
},
|
| 616 |
+
{
|
| 617 |
+
"epoch": 1.63,
|
| 618 |
+
"eval_distillation_accuracy_counterfactual": 0.5312981015905592,
|
| 619 |
+
"eval_distillation_accuracy_factual": 0.8055413032324269,
|
| 620 |
+
"eval_distillation_f1_counterfactual": 0.4617410563594155,
|
| 621 |
+
"eval_distillation_f1_factual": 0.7931325749560401,
|
| 622 |
+
"eval_groundtruth_accuracy_counterfactual": 0.5194971780400205,
|
| 623 |
+
"eval_groundtruth_f1_counterfactual": 0.4513515695315669,
|
| 624 |
+
"eval_groundtruth_f1_factual": 0.7019030914005773,
|
| 625 |
+
"eval_icace_cosine": 0.45170876383781433,
|
| 626 |
+
"eval_icace_l2": 0.6569298505783081,
|
| 627 |
+
"eval_icace_normdiff": 0.41902977228164673,
|
| 628 |
+
"eval_loss": 8.250597953796387,
|
| 629 |
+
"eval_runtime": 12.89,
|
| 630 |
+
"eval_samples_per_second": 302.405,
|
| 631 |
+
"eval_steps_per_second": 2.405,
|
| 632 |
+
"step": 950
|
| 633 |
+
},
|
| 634 |
+
{
|
| 635 |
+
"epoch": 1.65,
|
| 636 |
+
"learning_rate": 3.960138648180243e-05,
|
| 637 |
+
"loss": 8.4944,
|
| 638 |
+
"step": 960
|
| 639 |
+
},
|
| 640 |
+
{
|
| 641 |
+
"epoch": 1.68,
|
| 642 |
+
"learning_rate": 3.938474870017331e-05,
|
| 643 |
+
"loss": 8.4011,
|
| 644 |
+
"step": 980
|
| 645 |
+
},
|
| 646 |
+
{
|
| 647 |
+
"epoch": 1.72,
|
| 648 |
+
"learning_rate": 3.91681109185442e-05,
|
| 649 |
+
"loss": 8.4287,
|
| 650 |
+
"step": 1000
|
| 651 |
+
},
|
| 652 |
+
{
|
| 653 |
+
"epoch": 1.72,
|
| 654 |
+
"eval_distillation_accuracy_counterfactual": 0.5371985633658286,
|
| 655 |
+
"eval_distillation_accuracy_factual": 0.7909184197024115,
|
| 656 |
+
"eval_distillation_f1_counterfactual": 0.46983763588777594,
|
| 657 |
+
"eval_distillation_f1_factual": 0.7799708268449532,
|
| 658 |
+
"eval_groundtruth_accuracy_counterfactual": 0.5318111852231914,
|
| 659 |
+
"eval_groundtruth_f1_counterfactual": 0.4659493577740304,
|
| 660 |
+
"eval_groundtruth_f1_factual": 0.681861236036038,
|
| 661 |
+
"eval_icace_cosine": 0.4572613537311554,
|
| 662 |
+
"eval_icace_l2": 0.6384456753730774,
|
| 663 |
+
"eval_icace_normdiff": 0.3894091844558716,
|
| 664 |
+
"eval_loss": 8.310347557067871,
|
| 665 |
+
"eval_runtime": 12.9611,
|
| 666 |
+
"eval_samples_per_second": 300.747,
|
| 667 |
+
"eval_steps_per_second": 2.392,
|
| 668 |
+
"step": 1000
|
| 669 |
+
},
|
| 670 |
+
{
|
| 671 |
+
"epoch": 1.75,
|
| 672 |
+
"learning_rate": 3.895147313691508e-05,
|
| 673 |
+
"loss": 8.2598,
|
| 674 |
+
"step": 1020
|
| 675 |
+
},
|
| 676 |
+
{
|
| 677 |
+
"epoch": 1.78,
|
| 678 |
+
"learning_rate": 3.873483535528597e-05,
|
| 679 |
+
"loss": 8.2625,
|
| 680 |
+
"step": 1040
|
| 681 |
+
},
|
| 682 |
+
{
|
| 683 |
+
"epoch": 1.8,
|
| 684 |
+
"eval_distillation_accuracy_counterfactual": 0.5397639815289892,
|
| 685 |
+
"eval_distillation_accuracy_factual": 0.7875833760903027,
|
| 686 |
+
"eval_distillation_f1_counterfactual": 0.4799542696533782,
|
| 687 |
+
"eval_distillation_f1_factual": 0.7726528869168954,
|
| 688 |
+
"eval_groundtruth_accuracy_counterfactual": 0.5369420215495125,
|
| 689 |
+
"eval_groundtruth_f1_counterfactual": 0.4771898703044782,
|
| 690 |
+
"eval_groundtruth_f1_factual": 0.7023731499922957,
|
| 691 |
+
"eval_icace_cosine": 0.45770639181137085,
|
| 692 |
+
"eval_icace_l2": 0.6276518702507019,
|
| 693 |
+
"eval_icace_normdiff": 0.3755152225494385,
|
| 694 |
+
"eval_loss": 8.230777740478516,
|
| 695 |
+
"eval_runtime": 12.9942,
|
| 696 |
+
"eval_samples_per_second": 299.981,
|
| 697 |
+
"eval_steps_per_second": 2.386,
|
| 698 |
+
"step": 1050
|
| 699 |
+
},
|
| 700 |
+
{
|
| 701 |
+
"epoch": 1.82,
|
| 702 |
+
"learning_rate": 3.851819757365685e-05,
|
| 703 |
+
"loss": 8.0979,
|
| 704 |
+
"step": 1060
|
| 705 |
+
},
|
| 706 |
+
{
|
| 707 |
+
"epoch": 1.85,
|
| 708 |
+
"learning_rate": 3.830155979202773e-05,
|
| 709 |
+
"loss": 8.2971,
|
| 710 |
+
"step": 1080
|
| 711 |
+
},
|
| 712 |
+
{
|
| 713 |
+
"epoch": 1.89,
|
| 714 |
+
"learning_rate": 3.8084922010398616e-05,
|
| 715 |
+
"loss": 8.343,
|
| 716 |
+
"step": 1100
|
| 717 |
+
},
|
| 718 |
+
{
|
| 719 |
+
"epoch": 1.89,
|
| 720 |
+
"eval_distillation_accuracy_counterfactual": 0.5415597742432017,
|
| 721 |
+
"eval_distillation_accuracy_factual": 0.806054386865059,
|
| 722 |
+
"eval_distillation_f1_counterfactual": 0.4594314006577007,
|
| 723 |
+
"eval_distillation_f1_factual": 0.7921003157959313,
|
| 724 |
+
"eval_groundtruth_accuracy_counterfactual": 0.5400205233453053,
|
| 725 |
+
"eval_groundtruth_f1_counterfactual": 0.45903598954343383,
|
| 726 |
+
"eval_groundtruth_f1_factual": 0.6999320297801038,
|
| 727 |
+
"eval_icace_cosine": 0.45384570956230164,
|
| 728 |
+
"eval_icace_l2": 0.6284541487693787,
|
| 729 |
+
"eval_icace_normdiff": 0.3843493163585663,
|
| 730 |
+
"eval_loss": 8.147943496704102,
|
| 731 |
+
"eval_runtime": 13.1714,
|
| 732 |
+
"eval_samples_per_second": 295.944,
|
| 733 |
+
"eval_steps_per_second": 2.354,
|
| 734 |
+
"step": 1100
|
| 735 |
+
},
|
| 736 |
+
{
|
| 737 |
+
"epoch": 1.92,
|
| 738 |
+
"learning_rate": 3.78682842287695e-05,
|
| 739 |
+
"loss": 8.4675,
|
| 740 |
+
"step": 1120
|
| 741 |
+
},
|
| 742 |
+
{
|
| 743 |
+
"epoch": 1.96,
|
| 744 |
+
"learning_rate": 3.765164644714038e-05,
|
| 745 |
+
"loss": 8.3651,
|
| 746 |
+
"step": 1140
|
| 747 |
+
},
|
| 748 |
+
{
|
| 749 |
+
"epoch": 1.97,
|
| 750 |
+
"eval_distillation_accuracy_counterfactual": 0.5330938943047717,
|
| 751 |
+
"eval_distillation_accuracy_factual": 0.8273473576192919,
|
| 752 |
+
"eval_distillation_f1_counterfactual": 0.4438315486422179,
|
| 753 |
+
"eval_distillation_f1_factual": 0.8114092211560717,
|
| 754 |
+
"eval_groundtruth_accuracy_counterfactual": 0.5282195997947665,
|
| 755 |
+
"eval_groundtruth_f1_counterfactual": 0.43912524130185177,
|
| 756 |
+
"eval_groundtruth_f1_factual": 0.696958037051927,
|
| 757 |
+
"eval_icace_cosine": 0.45857784152030945,
|
| 758 |
+
"eval_icace_l2": 0.6330906748771667,
|
| 759 |
+
"eval_icace_normdiff": 0.3742936849594116,
|
| 760 |
+
"eval_loss": 8.261628150939941,
|
| 761 |
+
"eval_runtime": 12.8921,
|
| 762 |
+
"eval_samples_per_second": 302.355,
|
| 763 |
+
"eval_steps_per_second": 2.405,
|
| 764 |
+
"step": 1150
|
| 765 |
+
},
|
| 766 |
+
{
|
| 767 |
+
"epoch": 1.99,
|
| 768 |
+
"learning_rate": 3.7435008665511264e-05,
|
| 769 |
+
"loss": 8.3106,
|
| 770 |
+
"step": 1160
|
| 771 |
+
},
|
| 772 |
+
{
|
| 773 |
+
"epoch": 2.02,
|
| 774 |
+
"learning_rate": 3.7218370883882154e-05,
|
| 775 |
+
"loss": 8.1733,
|
| 776 |
+
"step": 1180
|
| 777 |
+
},
|
| 778 |
+
{
|
| 779 |
+
"epoch": 2.06,
|
| 780 |
+
"learning_rate": 3.700173310225303e-05,
|
| 781 |
+
"loss": 8.1566,
|
| 782 |
+
"step": 1200
|
| 783 |
+
},
|
| 784 |
+
{
|
| 785 |
+
"epoch": 2.06,
|
| 786 |
+
"eval_distillation_accuracy_counterfactual": 0.521806054386865,
|
| 787 |
+
"eval_distillation_accuracy_factual": 0.8083632632119035,
|
| 788 |
+
"eval_distillation_f1_counterfactual": 0.4568626686367353,
|
| 789 |
+
"eval_distillation_f1_factual": 0.793454263258986,
|
| 790 |
+
"eval_groundtruth_accuracy_counterfactual": 0.5210364289379169,
|
| 791 |
+
"eval_groundtruth_f1_counterfactual": 0.4589985732848542,
|
| 792 |
+
"eval_groundtruth_f1_factual": 0.6909411446988759,
|
| 793 |
+
"eval_icace_cosine": 0.4588005244731903,
|
| 794 |
+
"eval_icace_l2": 0.6523511409759521,
|
| 795 |
+
"eval_icace_normdiff": 0.40103352069854736,
|
| 796 |
+
"eval_loss": 8.481898307800293,
|
| 797 |
+
"eval_runtime": 12.9624,
|
| 798 |
+
"eval_samples_per_second": 300.717,
|
| 799 |
+
"eval_steps_per_second": 2.392,
|
| 800 |
+
"step": 1200
|
| 801 |
+
},
|
| 802 |
+
{
|
| 803 |
+
"epoch": 2.09,
|
| 804 |
+
"learning_rate": 3.678509532062392e-05,
|
| 805 |
+
"loss": 8.0229,
|
| 806 |
+
"step": 1220
|
| 807 |
+
},
|
| 808 |
+
{
|
| 809 |
+
"epoch": 2.13,
|
| 810 |
+
"learning_rate": 3.65684575389948e-05,
|
| 811 |
+
"loss": 8.3726,
|
| 812 |
+
"step": 1240
|
| 813 |
+
},
|
| 814 |
+
{
|
| 815 |
+
"epoch": 2.14,
|
| 816 |
+
"eval_distillation_accuracy_counterfactual": 0.5246280143663418,
|
| 817 |
+
"eval_distillation_accuracy_factual": 0.8040020523345305,
|
| 818 |
+
"eval_distillation_f1_counterfactual": 0.4638517088409774,
|
| 819 |
+
"eval_distillation_f1_factual": 0.7892649860700691,
|
| 820 |
+
"eval_groundtruth_accuracy_counterfactual": 0.5166752180605438,
|
| 821 |
+
"eval_groundtruth_f1_counterfactual": 0.4574432564220393,
|
| 822 |
+
"eval_groundtruth_f1_factual": 0.7129214655450473,
|
| 823 |
+
"eval_icace_cosine": 0.45950308442115784,
|
| 824 |
+
"eval_icace_l2": 0.6577529311180115,
|
| 825 |
+
"eval_icace_normdiff": 0.4075782299041748,
|
| 826 |
+
"eval_loss": 8.51949691772461,
|
| 827 |
+
"eval_runtime": 12.9172,
|
| 828 |
+
"eval_samples_per_second": 301.769,
|
| 829 |
+
"eval_steps_per_second": 2.4,
|
| 830 |
+
"step": 1250
|
| 831 |
+
},
|
| 832 |
+
{
|
| 833 |
+
"epoch": 2.16,
|
| 834 |
+
"learning_rate": 3.6351819757365686e-05,
|
| 835 |
+
"loss": 8.2033,
|
| 836 |
+
"step": 1260
|
| 837 |
+
},
|
| 838 |
+
{
|
| 839 |
+
"epoch": 2.2,
|
| 840 |
+
"learning_rate": 3.613518197573657e-05,
|
| 841 |
+
"loss": 8.2507,
|
| 842 |
+
"step": 1280
|
| 843 |
+
},
|
| 844 |
+
{
|
| 845 |
+
"epoch": 2.23,
|
| 846 |
+
"learning_rate": 3.591854419410746e-05,
|
| 847 |
+
"loss": 8.0753,
|
| 848 |
+
"step": 1300
|
| 849 |
+
},
|
| 850 |
+
{
|
| 851 |
+
"epoch": 2.23,
|
| 852 |
+
"eval_distillation_accuracy_counterfactual": 0.5397639815289892,
|
| 853 |
+
"eval_distillation_accuracy_factual": 0.8165726013340174,
|
| 854 |
+
"eval_distillation_f1_counterfactual": 0.47682727590749313,
|
| 855 |
+
"eval_distillation_f1_factual": 0.7967308018931318,
|
| 856 |
+
"eval_groundtruth_accuracy_counterfactual": 0.5325808106721396,
|
| 857 |
+
"eval_groundtruth_f1_counterfactual": 0.47018455613210597,
|
| 858 |
+
"eval_groundtruth_f1_factual": 0.7043239542576792,
|
| 859 |
+
"eval_icace_cosine": 0.4495893120765686,
|
| 860 |
+
"eval_icace_l2": 0.6213250160217285,
|
| 861 |
+
"eval_icace_normdiff": 0.37207138538360596,
|
| 862 |
+
"eval_loss": 8.11995792388916,
|
| 863 |
+
"eval_runtime": 13.9585,
|
| 864 |
+
"eval_samples_per_second": 279.256,
|
| 865 |
+
"eval_steps_per_second": 2.221,
|
| 866 |
+
"step": 1300
|
| 867 |
+
},
|
| 868 |
+
{
|
| 869 |
+
"epoch": 2.26,
|
| 870 |
+
"learning_rate": 3.5701906412478334e-05,
|
| 871 |
+
"loss": 8.2577,
|
| 872 |
+
"step": 1320
|
| 873 |
+
},
|
| 874 |
+
{
|
| 875 |
+
"epoch": 2.3,
|
| 876 |
+
"learning_rate": 3.5485268630849224e-05,
|
| 877 |
+
"loss": 8.3416,
|
| 878 |
+
"step": 1340
|
| 879 |
+
},
|
| 880 |
+
{
|
| 881 |
+
"epoch": 2.32,
|
| 882 |
+
"eval_distillation_accuracy_counterfactual": 0.5246280143663418,
|
| 883 |
+
"eval_distillation_accuracy_factual": 0.8183683940482298,
|
| 884 |
+
"eval_distillation_f1_counterfactual": 0.4426092418935421,
|
| 885 |
+
"eval_distillation_f1_factual": 0.8016386570803451,
|
| 886 |
+
"eval_groundtruth_accuracy_counterfactual": 0.5066700872242176,
|
| 887 |
+
"eval_groundtruth_f1_counterfactual": 0.43023988421850207,
|
| 888 |
+
"eval_groundtruth_f1_factual": 0.7132188999952972,
|
| 889 |
+
"eval_icace_cosine": 0.45355769991874695,
|
| 890 |
+
"eval_icace_l2": 0.6458914279937744,
|
| 891 |
+
"eval_icace_normdiff": 0.3992498517036438,
|
| 892 |
+
"eval_loss": 8.449216842651367,
|
| 893 |
+
"eval_runtime": 12.8687,
|
| 894 |
+
"eval_samples_per_second": 302.906,
|
| 895 |
+
"eval_steps_per_second": 2.409,
|
| 896 |
+
"step": 1350
|
| 897 |
+
},
|
| 898 |
+
{
|
| 899 |
+
"epoch": 2.33,
|
| 900 |
+
"learning_rate": 3.526863084922011e-05,
|
| 901 |
+
"loss": 8.2811,
|
| 902 |
+
"step": 1360
|
| 903 |
+
},
|
| 904 |
+
{
|
| 905 |
+
"epoch": 2.37,
|
| 906 |
+
"learning_rate": 3.505199306759099e-05,
|
| 907 |
+
"loss": 8.192,
|
| 908 |
+
"step": 1380
|
| 909 |
+
},
|
| 910 |
+
{
|
| 911 |
+
"epoch": 2.4,
|
| 912 |
+
"learning_rate": 3.483535528596187e-05,
|
| 913 |
+
"loss": 8.1159,
|
| 914 |
+
"step": 1400
|
| 915 |
+
},
|
| 916 |
+
{
|
| 917 |
+
"epoch": 2.4,
|
| 918 |
+
"eval_distillation_accuracy_counterfactual": 0.5348896870189841,
|
| 919 |
+
"eval_distillation_accuracy_factual": 0.8070805541303232,
|
| 920 |
+
"eval_distillation_f1_counterfactual": 0.4442360778869393,
|
| 921 |
+
"eval_distillation_f1_factual": 0.793089455272879,
|
| 922 |
+
"eval_groundtruth_accuracy_counterfactual": 0.5389943560800411,
|
| 923 |
+
"eval_groundtruth_f1_counterfactual": 0.45311551419493207,
|
| 924 |
+
"eval_groundtruth_f1_factual": 0.713127845360182,
|
| 925 |
+
"eval_icace_cosine": 0.45396122336387634,
|
| 926 |
+
"eval_icace_l2": 0.6456016302108765,
|
| 927 |
+
"eval_icace_normdiff": 0.397720605134964,
|
| 928 |
+
"eval_loss": 8.26937198638916,
|
| 929 |
+
"eval_runtime": 13.0046,
|
| 930 |
+
"eval_samples_per_second": 299.74,
|
| 931 |
+
"eval_steps_per_second": 2.384,
|
| 932 |
+
"step": 1400
|
| 933 |
+
},
|
| 934 |
+
{
|
| 935 |
+
"epoch": 2.44,
|
| 936 |
+
"learning_rate": 3.461871750433276e-05,
|
| 937 |
+
"loss": 8.1856,
|
| 938 |
+
"step": 1420
|
| 939 |
+
},
|
| 940 |
+
{
|
| 941 |
+
"epoch": 2.47,
|
| 942 |
+
"learning_rate": 3.440207972270364e-05,
|
| 943 |
+
"loss": 8.2343,
|
| 944 |
+
"step": 1440
|
| 945 |
+
},
|
| 946 |
+
{
|
| 947 |
+
"epoch": 2.49,
|
| 948 |
+
"eval_distillation_accuracy_counterfactual": 0.5105182144689584,
|
| 949 |
+
"eval_distillation_accuracy_factual": 0.8150333504361211,
|
| 950 |
+
"eval_distillation_f1_counterfactual": 0.46519275113167036,
|
| 951 |
+
"eval_distillation_f1_factual": 0.7986983661724563,
|
| 952 |
+
"eval_groundtruth_accuracy_counterfactual": 0.5051308363263212,
|
| 953 |
+
"eval_groundtruth_f1_counterfactual": 0.4604440377605913,
|
| 954 |
+
"eval_groundtruth_f1_factual": 0.6976730796070983,
|
| 955 |
+
"eval_icace_cosine": 0.44814541935920715,
|
| 956 |
+
"eval_icace_l2": 0.6517916917800903,
|
| 957 |
+
"eval_icace_normdiff": 0.4058307409286499,
|
| 958 |
+
"eval_loss": 8.407746315002441,
|
| 959 |
+
"eval_runtime": 12.9229,
|
| 960 |
+
"eval_samples_per_second": 301.635,
|
| 961 |
+
"eval_steps_per_second": 2.399,
|
| 962 |
+
"step": 1450
|
| 963 |
+
}
|
| 964 |
+
],
|
| 965 |
+
"max_steps": 4616,
|
| 966 |
+
"num_train_epochs": 8,
|
| 967 |
+
"total_flos": 1.2202103856768e+16,
|
| 968 |
+
"trial_name": null,
|
| 969 |
+
"trial_params": null
|
| 970 |
+
}
|
training_args.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:280559b176d6b208233ebbac8d6b4100390edc06b74dde51155ce210d0105a5e
|
| 3 |
+
size 3183
|
vocab.json
ADDED
|
The diff for this file is too large to render.
See raw diff
|
|
|