Upload folder using huggingface_hub
Browse files- checkpoint-100/config.json +41 -0
- checkpoint-100/model.safetensors +3 -0
- checkpoint-100/rng_state.pth +3 -0
- checkpoint-100/training_args.bin +3 -0
- checkpoint-225/model.safetensors +1 -1
- choice_distribution.json +7 -0
- logs/events.out.tfevents.1754579008.209-20-159-47.43249.0 +3 -0
- logs/events.out.tfevents.1754579008.209-20-159-47.43249.1 +3 -0
- model.safetensors +1 -1
- per_sample_predictions.csv +81 -0
checkpoint-100/config.json
ADDED
|
@@ -0,0 +1,41 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"architectures": [
|
| 3 |
+
"RobertaForSequenceClassification"
|
| 4 |
+
],
|
| 5 |
+
"attention_probs_dropout_prob": 0.1,
|
| 6 |
+
"bos_token_id": 0,
|
| 7 |
+
"classifier_dropout": null,
|
| 8 |
+
"eos_token_id": 2,
|
| 9 |
+
"hidden_act": "gelu",
|
| 10 |
+
"hidden_dropout_prob": 0.1,
|
| 11 |
+
"hidden_size": 768,
|
| 12 |
+
"id2label": {
|
| 13 |
+
"0": "LABEL_0",
|
| 14 |
+
"1": "LABEL_1",
|
| 15 |
+
"2": "LABEL_2",
|
| 16 |
+
"3": "LABEL_3",
|
| 17 |
+
"4": "LABEL_4"
|
| 18 |
+
},
|
| 19 |
+
"initializer_range": 0.02,
|
| 20 |
+
"intermediate_size": 3072,
|
| 21 |
+
"label2id": {
|
| 22 |
+
"LABEL_0": 0,
|
| 23 |
+
"LABEL_1": 1,
|
| 24 |
+
"LABEL_2": 2,
|
| 25 |
+
"LABEL_3": 3,
|
| 26 |
+
"LABEL_4": 4
|
| 27 |
+
},
|
| 28 |
+
"layer_norm_eps": 1e-05,
|
| 29 |
+
"max_position_embeddings": 514,
|
| 30 |
+
"model_type": "roberta",
|
| 31 |
+
"num_attention_heads": 12,
|
| 32 |
+
"num_hidden_layers": 12,
|
| 33 |
+
"pad_token_id": 1,
|
| 34 |
+
"position_embedding_type": "absolute",
|
| 35 |
+
"problem_type": "multi_label_classification",
|
| 36 |
+
"torch_dtype": "bfloat16",
|
| 37 |
+
"transformers_version": "4.55.0",
|
| 38 |
+
"type_vocab_size": 1,
|
| 39 |
+
"use_cache": false,
|
| 40 |
+
"vocab_size": 50265
|
| 41 |
+
}
|
checkpoint-100/model.safetensors
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:5389e4740b349a1cb2c5c4439848d3de1f4d0e7ccf00c3eba3383a4706ade0b1
|
| 3 |
+
size 249323242
|
checkpoint-100/rng_state.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:fb6684c2f3386c6e7cc69e9bc23875f477a31b9f1fc9457d83127b5c896ece7b
|
| 3 |
+
size 14645
|
checkpoint-100/training_args.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:4e660e5834bae29786f51d10e9e11cd4547ee3fc8a92996af55b6d5850bc1c8e
|
| 3 |
+
size 5713
|
checkpoint-225/model.safetensors
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 249323242
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:345bbc11175454ee2ed696c00112eb6377b190efa76ea62139fc29c9e9657210
|
| 3 |
size 249323242
|
choice_distribution.json
ADDED
|
@@ -0,0 +1,7 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"gpt-4o": 0.0,
|
| 3 |
+
"DeepSeek-V3-0324": 0.125,
|
| 4 |
+
"Llama-4-maverick-17b-128e-instruct-fp8": 0.0,
|
| 5 |
+
"qwen25-coder-32b-instruct": 0.0875,
|
| 6 |
+
"gpt-4.1-mini": 0.7875
|
| 7 |
+
}
|
logs/events.out.tfevents.1754579008.209-20-159-47.43249.0
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:a3b34f911f0622746bebe0da273ce3f8be17ea847538df8152b204ed6c33d525
|
| 3 |
+
size 7091
|
logs/events.out.tfevents.1754579008.209-20-159-47.43249.1
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:d0aa8b37c9a609ddf3124faaa499796840719a4a352bc6a93bd6a0b96c5ef332
|
| 3 |
+
size 7091
|
model.safetensors
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 249323242
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:5389e4740b349a1cb2c5c4439848d3de1f4d0e7ccf00c3eba3383a4706ade0b1
|
| 3 |
size 249323242
|
per_sample_predictions.csv
ADDED
|
@@ -0,0 +1,81 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
gpt-4o,DeepSeek-V3-0324,Llama-4-maverick-17b-128e-instruct-fp8,qwen25-coder-32b-instruct,gpt-4.1-mini,chosen_executor
|
| 2 |
+
-0.7734375,-0.66796875,-0.71875,-0.6796875,-0.58984375,gpt-4.1-mini
|
| 3 |
+
-0.890625,-0.765625,-0.87109375,-0.765625,-0.75,gpt-4.1-mini
|
| 4 |
+
-0.95703125,-0.8125,-0.87109375,-0.7890625,-0.74609375,gpt-4.1-mini
|
| 5 |
+
-0.91796875,-0.77734375,-0.875,-0.8125,-0.77734375,DeepSeek-V3-0324
|
| 6 |
+
-0.90625,-0.78125,-0.83203125,-0.7734375,-0.74609375,gpt-4.1-mini
|
| 7 |
+
-0.890625,-0.7734375,-0.859375,-0.78125,-0.75390625,gpt-4.1-mini
|
| 8 |
+
-0.8984375,-0.7578125,-0.80859375,-0.73828125,-0.67578125,gpt-4.1-mini
|
| 9 |
+
-0.89453125,-0.76953125,-0.83984375,-0.78515625,-0.75,gpt-4.1-mini
|
| 10 |
+
-0.78515625,-0.71875,-0.70703125,-0.6796875,-0.625,gpt-4.1-mini
|
| 11 |
+
-0.87890625,-0.77734375,-0.8359375,-0.73046875,-0.75390625,qwen25-coder-32b-instruct
|
| 12 |
+
-0.95703125,-0.8125,-0.8828125,-0.81640625,-0.8046875,gpt-4.1-mini
|
| 13 |
+
-0.8515625,-0.76171875,-0.83984375,-0.78515625,-0.7265625,gpt-4.1-mini
|
| 14 |
+
-0.93359375,-0.84375,-0.8671875,-0.79296875,-0.8125,qwen25-coder-32b-instruct
|
| 15 |
+
-0.875,-0.75,-0.79296875,-0.74609375,-0.69921875,gpt-4.1-mini
|
| 16 |
+
-0.9453125,-0.875,-0.8828125,-0.8046875,-0.796875,gpt-4.1-mini
|
| 17 |
+
-0.84765625,-0.76953125,-0.80078125,-0.7578125,-0.734375,gpt-4.1-mini
|
| 18 |
+
-0.9375,-0.78515625,-0.8671875,-0.79296875,-0.75390625,gpt-4.1-mini
|
| 19 |
+
-0.89453125,-0.75390625,-0.80859375,-0.78515625,-0.7265625,gpt-4.1-mini
|
| 20 |
+
-0.9609375,-0.84375,-0.9140625,-0.8125,-0.8125,qwen25-coder-32b-instruct
|
| 21 |
+
-0.91015625,-0.8125,-0.875,-0.81640625,-0.7421875,gpt-4.1-mini
|
| 22 |
+
-0.921875,-0.82421875,-0.9140625,-0.79296875,-0.7734375,gpt-4.1-mini
|
| 23 |
+
-0.8515625,-0.734375,-0.796875,-0.71484375,-0.6640625,gpt-4.1-mini
|
| 24 |
+
-0.984375,-0.82421875,-0.87890625,-0.83203125,-0.76171875,gpt-4.1-mini
|
| 25 |
+
-0.8984375,-0.83203125,-0.87109375,-0.8125,-0.78125,gpt-4.1-mini
|
| 26 |
+
-0.94140625,-0.81640625,-0.91796875,-0.8515625,-0.83203125,DeepSeek-V3-0324
|
| 27 |
+
-0.87890625,-0.6796875,-0.72265625,-0.7265625,-0.65234375,gpt-4.1-mini
|
| 28 |
+
-0.8984375,-0.796875,-0.859375,-0.78125,-0.7578125,gpt-4.1-mini
|
| 29 |
+
-0.90234375,-0.77734375,-0.88671875,-0.796875,-0.7734375,gpt-4.1-mini
|
| 30 |
+
-0.9296875,-0.78515625,-0.8828125,-0.79296875,-0.75390625,gpt-4.1-mini
|
| 31 |
+
-0.83984375,-0.71875,-0.80859375,-0.7578125,-0.6875,gpt-4.1-mini
|
| 32 |
+
-0.8671875,-0.76953125,-0.84375,-0.76171875,-0.73828125,gpt-4.1-mini
|
| 33 |
+
-0.87109375,-0.7734375,-0.828125,-0.76171875,-0.75390625,gpt-4.1-mini
|
| 34 |
+
-0.8671875,-0.71484375,-0.796875,-0.71484375,-0.6875,gpt-4.1-mini
|
| 35 |
+
-0.8828125,-0.7734375,-0.83203125,-0.76171875,-0.73828125,gpt-4.1-mini
|
| 36 |
+
-0.9765625,-0.83203125,-0.875,-0.83203125,-0.78125,gpt-4.1-mini
|
| 37 |
+
-0.90625,-0.796875,-0.8828125,-0.78515625,-0.796875,qwen25-coder-32b-instruct
|
| 38 |
+
-0.9140625,-0.81640625,-0.859375,-0.80078125,-0.74609375,gpt-4.1-mini
|
| 39 |
+
-0.91796875,-0.80078125,-0.87109375,-0.8046875,-0.7734375,gpt-4.1-mini
|
| 40 |
+
-0.90625,-0.80859375,-0.8359375,-0.79296875,-0.75390625,gpt-4.1-mini
|
| 41 |
+
-0.921875,-0.796875,-0.86328125,-0.76171875,-0.76953125,qwen25-coder-32b-instruct
|
| 42 |
+
-0.9296875,-0.78125,-0.8671875,-0.76953125,-0.74609375,gpt-4.1-mini
|
| 43 |
+
-0.90625,-0.7890625,-0.8671875,-0.7890625,-0.765625,gpt-4.1-mini
|
| 44 |
+
-0.9609375,-0.84765625,-0.89453125,-0.80859375,-0.734375,gpt-4.1-mini
|
| 45 |
+
-0.9375,-0.84375,-0.90234375,-0.84375,-0.796875,gpt-4.1-mini
|
| 46 |
+
-0.9296875,-0.796875,-0.8984375,-0.81640625,-0.8125,DeepSeek-V3-0324
|
| 47 |
+
-0.87890625,-0.75,-0.8359375,-0.76171875,-0.7421875,gpt-4.1-mini
|
| 48 |
+
-0.8671875,-0.7578125,-0.828125,-0.76171875,-0.71484375,gpt-4.1-mini
|
| 49 |
+
-0.96484375,-0.8359375,-0.86328125,-0.80078125,-0.73828125,gpt-4.1-mini
|
| 50 |
+
-0.98046875,-0.8671875,-0.91796875,-0.84375,-0.80078125,gpt-4.1-mini
|
| 51 |
+
-0.8984375,-0.80078125,-0.84375,-0.7734375,-0.7578125,gpt-4.1-mini
|
| 52 |
+
-0.8984375,-0.73046875,-0.85546875,-0.77734375,-0.765625,DeepSeek-V3-0324
|
| 53 |
+
-0.875,-0.7578125,-0.83984375,-0.80078125,-0.68359375,gpt-4.1-mini
|
| 54 |
+
-0.92578125,-0.8046875,-0.8359375,-0.77734375,-0.7421875,gpt-4.1-mini
|
| 55 |
+
-0.8125,-0.66796875,-0.7421875,-0.66015625,-0.625,gpt-4.1-mini
|
| 56 |
+
-0.9609375,-0.81640625,-0.875,-0.8046875,-0.7734375,gpt-4.1-mini
|
| 57 |
+
-0.9375,-0.87890625,-0.921875,-0.8125,-0.78515625,gpt-4.1-mini
|
| 58 |
+
-0.9375,-0.8203125,-0.8984375,-0.8359375,-0.77734375,gpt-4.1-mini
|
| 59 |
+
-0.859375,-0.72265625,-0.80078125,-0.7109375,-0.671875,gpt-4.1-mini
|
| 60 |
+
-0.87109375,-0.75,-0.8515625,-0.7890625,-0.73828125,gpt-4.1-mini
|
| 61 |
+
-0.9453125,-0.828125,-0.9140625,-0.8203125,-0.8203125,qwen25-coder-32b-instruct
|
| 62 |
+
-0.95703125,-0.81640625,-0.90234375,-0.84765625,-0.78125,gpt-4.1-mini
|
| 63 |
+
-0.91796875,-0.83984375,-0.8515625,-0.8125,-0.76171875,gpt-4.1-mini
|
| 64 |
+
-0.93359375,-0.8203125,-0.91796875,-0.81640625,-0.81640625,qwen25-coder-32b-instruct
|
| 65 |
+
-0.8828125,-0.76953125,-0.8359375,-0.78125,-0.76953125,DeepSeek-V3-0324
|
| 66 |
+
-0.859375,-0.73828125,-0.8359375,-0.76171875,-0.74609375,DeepSeek-V3-0324
|
| 67 |
+
-0.82421875,-0.72265625,-0.80078125,-0.7265625,-0.6953125,gpt-4.1-mini
|
| 68 |
+
-0.87890625,-0.7109375,-0.8515625,-0.7734375,-0.73046875,DeepSeek-V3-0324
|
| 69 |
+
-0.8828125,-0.765625,-0.8515625,-0.74609375,-0.734375,gpt-4.1-mini
|
| 70 |
+
-0.84765625,-0.6953125,-0.80078125,-0.73828125,-0.7109375,DeepSeek-V3-0324
|
| 71 |
+
-0.9140625,-0.828125,-0.859375,-0.79296875,-0.75390625,gpt-4.1-mini
|
| 72 |
+
-0.9140625,-0.77734375,-0.8671875,-0.80078125,-0.765625,gpt-4.1-mini
|
| 73 |
+
-0.890625,-0.8046875,-0.828125,-0.75,-0.734375,gpt-4.1-mini
|
| 74 |
+
-0.8984375,-0.7578125,-0.859375,-0.76953125,-0.7734375,DeepSeek-V3-0324
|
| 75 |
+
-0.91796875,-0.83984375,-0.8828125,-0.7890625,-0.78125,gpt-4.1-mini
|
| 76 |
+
-0.93359375,-0.8046875,-0.86328125,-0.80859375,-0.7578125,gpt-4.1-mini
|
| 77 |
+
-0.9296875,-0.77734375,-0.890625,-0.81640625,-0.80859375,DeepSeek-V3-0324
|
| 78 |
+
-0.83984375,-0.7265625,-0.8515625,-0.75,-0.6953125,gpt-4.1-mini
|
| 79 |
+
-0.9375,-0.82421875,-0.9140625,-0.82421875,-0.81640625,gpt-4.1-mini
|
| 80 |
+
-0.9609375,-0.84375,-0.90234375,-0.83203125,-0.80078125,gpt-4.1-mini
|
| 81 |
+
-0.89453125,-0.80078125,-0.875,-0.81640625,-0.796875,gpt-4.1-mini
|