Model save
Browse files- README.md +2 -3
- all_results.json +4 -5
- config.json +1 -2
- generation_config.json +1 -1
- model-00001-of-00003.safetensors +1 -1
- model-00002-of-00003.safetensors +1 -1
- model-00003-of-00003.safetensors +1 -1
- tokenizer.json +0 -1
- tokenizer_config.json +1 -2
- train_results.json +4 -5
- trainer_state.json +60 -98
- training_args.bin +2 -2
README.md
CHANGED
@@ -13,7 +13,6 @@ model-index:
|
|
13 |
<!-- This model card has been generated automatically according to the information the Trainer had access to. You
|
14 |
should probably proofread and complete it, then remove this comment. -->
|
15 |
|
16 |
-
[<img src="https://raw.githubusercontent.com/wandb/assets/main/wandb-github-badge-28.svg" alt="Visualize in Weights & Biases" width="200" height="32"/>](https://wandb.ai/sanqiang/wdpo/runs/h1ajp915)
|
17 |
# zephyr-7b-dpo-full
|
18 |
|
19 |
This model is a fine-tuned version of [HuggingFaceH4/mistral-7b-sft-beta](https://huggingface.co/HuggingFaceH4/mistral-7b-sft-beta) on the None dataset.
|
@@ -55,7 +54,7 @@ The following hyperparameters were used during training:
|
|
55 |
|
56 |
### Framework versions
|
57 |
|
58 |
-
- Transformers 4.
|
59 |
- Pytorch 2.1.2+cu121
|
60 |
- Datasets 2.14.6
|
61 |
-
- Tokenizers 0.
|
|
|
13 |
<!-- This model card has been generated automatically according to the information the Trainer had access to. You
|
14 |
should probably proofread and complete it, then remove this comment. -->
|
15 |
|
|
|
16 |
# zephyr-7b-dpo-full
|
17 |
|
18 |
This model is a fine-tuned version of [HuggingFaceH4/mistral-7b-sft-beta](https://huggingface.co/HuggingFaceH4/mistral-7b-sft-beta) on the None dataset.
|
|
|
54 |
|
55 |
### Framework versions
|
56 |
|
57 |
+
- Transformers 4.35.2
|
58 |
- Pytorch 2.1.2+cu121
|
59 |
- Datasets 2.14.6
|
60 |
+
- Tokenizers 0.14.1
|
all_results.json
CHANGED
@@ -1,9 +1,8 @@
|
|
1 |
{
|
2 |
"epoch": 1.0,
|
3 |
-
"
|
4 |
-
"
|
5 |
-
"train_runtime": 385.1956,
|
6 |
"train_samples": 6750,
|
7 |
-
"train_samples_per_second":
|
8 |
-
"train_steps_per_second": 0.
|
9 |
}
|
|
|
1 |
{
|
2 |
"epoch": 1.0,
|
3 |
+
"train_loss": 0.6385756753525644,
|
4 |
+
"train_runtime": 422.6241,
|
|
|
5 |
"train_samples": 6750,
|
6 |
+
"train_samples_per_second": 15.972,
|
7 |
+
"train_steps_per_second": 0.125
|
8 |
}
|
config.json
CHANGED
@@ -3,7 +3,6 @@
|
|
3 |
"architectures": [
|
4 |
"MistralForCausalLM"
|
5 |
],
|
6 |
-
"attention_dropout": 0.0,
|
7 |
"bos_token_id": 1,
|
8 |
"eos_token_id": 2,
|
9 |
"hidden_act": "silu",
|
@@ -20,7 +19,7 @@
|
|
20 |
"sliding_window": 4096,
|
21 |
"tie_word_embeddings": false,
|
22 |
"torch_dtype": "bfloat16",
|
23 |
-
"transformers_version": "4.
|
24 |
"use_cache": false,
|
25 |
"vocab_size": 32000
|
26 |
}
|
|
|
3 |
"architectures": [
|
4 |
"MistralForCausalLM"
|
5 |
],
|
|
|
6 |
"bos_token_id": 1,
|
7 |
"eos_token_id": 2,
|
8 |
"hidden_act": "silu",
|
|
|
19 |
"sliding_window": 4096,
|
20 |
"tie_word_embeddings": false,
|
21 |
"torch_dtype": "bfloat16",
|
22 |
+
"transformers_version": "4.35.2",
|
23 |
"use_cache": false,
|
24 |
"vocab_size": 32000
|
25 |
}
|
generation_config.json
CHANGED
@@ -2,5 +2,5 @@
|
|
2 |
"_from_model_config": true,
|
3 |
"bos_token_id": 1,
|
4 |
"eos_token_id": 2,
|
5 |
-
"transformers_version": "4.
|
6 |
}
|
|
|
2 |
"_from_model_config": true,
|
3 |
"bos_token_id": 1,
|
4 |
"eos_token_id": 2,
|
5 |
+
"transformers_version": "4.35.2"
|
6 |
}
|
model-00001-of-00003.safetensors
CHANGED
@@ -1,3 +1,3 @@
|
|
1 |
version https://git-lfs.github.com/spec/v1
|
2 |
-
oid sha256:
|
3 |
size 4943162336
|
|
|
1 |
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:6c5eccf170cb2807dafed373e6970a49edaa5ad3f88e5e35e03e143abba6dcb2
|
3 |
size 4943162336
|
model-00002-of-00003.safetensors
CHANGED
@@ -1,3 +1,3 @@
|
|
1 |
version https://git-lfs.github.com/spec/v1
|
2 |
-
oid sha256:
|
3 |
size 4999819336
|
|
|
1 |
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:b7cb9a2089e1f16a8ba2daaa3b1a78b5e41b35c6b1086cfa84a7c364f26c2418
|
3 |
size 4999819336
|
model-00003-of-00003.safetensors
CHANGED
@@ -1,3 +1,3 @@
|
|
1 |
version https://git-lfs.github.com/spec/v1
|
2 |
-
oid sha256:
|
3 |
size 4540516344
|
|
|
1 |
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:a44cd1f203562bd6cb7aff11086c30c740baefe049f30b9c91acb227c1de6938
|
3 |
size 4540516344
|
tokenizer.json
CHANGED
@@ -134,7 +134,6 @@
|
|
134 |
"end_of_word_suffix": null,
|
135 |
"fuse_unk": true,
|
136 |
"byte_fallback": true,
|
137 |
-
"ignore_merges": false,
|
138 |
"vocab": {
|
139 |
"<unk>": 0,
|
140 |
"<s>": 1,
|
|
|
134 |
"end_of_word_suffix": null,
|
135 |
"fuse_unk": true,
|
136 |
"byte_fallback": true,
|
|
|
137 |
"vocab": {
|
138 |
"<unk>": 0,
|
139 |
"<s>": 1,
|
tokenizer_config.json
CHANGED
@@ -1,6 +1,4 @@
|
|
1 |
{
|
2 |
-
"add_bos_token": true,
|
3 |
-
"add_eos_token": false,
|
4 |
"added_tokens_decoder": {
|
5 |
"0": {
|
6 |
"content": "<unk>",
|
@@ -36,6 +34,7 @@
|
|
36 |
"chat_template": "{% for message in messages %}\n{% if message['role'] == 'user' %}\n{{ '<|user|>\n' + message['content'] + eos_token }}\n{% elif message['role'] == 'system' %}\n{{ '<|system|>\n' + message['content'] + eos_token }}\n{% elif message['role'] == 'assistant' %}\n{{ '<|assistant|>\n' + message['content'] + eos_token }}\n{% endif %}\n{% if loop.last and add_generation_prompt %}\n{{ '<|assistant|>' }}\n{% endif %}\n{% endfor %}",
|
37 |
"clean_up_tokenization_spaces": false,
|
38 |
"eos_token": "</s>",
|
|
|
39 |
"model_max_length": 2048,
|
40 |
"pad_token": "</s>",
|
41 |
"sp_model_kwargs": {},
|
|
|
1 |
{
|
|
|
|
|
2 |
"added_tokens_decoder": {
|
3 |
"0": {
|
4 |
"content": "<unk>",
|
|
|
34 |
"chat_template": "{% for message in messages %}\n{% if message['role'] == 'user' %}\n{{ '<|user|>\n' + message['content'] + eos_token }}\n{% elif message['role'] == 'system' %}\n{{ '<|system|>\n' + message['content'] + eos_token }}\n{% elif message['role'] == 'assistant' %}\n{{ '<|assistant|>\n' + message['content'] + eos_token }}\n{% endif %}\n{% if loop.last and add_generation_prompt %}\n{{ '<|assistant|>' }}\n{% endif %}\n{% endfor %}",
|
35 |
"clean_up_tokenization_spaces": false,
|
36 |
"eos_token": "</s>",
|
37 |
+
"legacy": true,
|
38 |
"model_max_length": 2048,
|
39 |
"pad_token": "</s>",
|
40 |
"sp_model_kwargs": {},
|
train_results.json
CHANGED
@@ -1,9 +1,8 @@
|
|
1 |
{
|
2 |
"epoch": 1.0,
|
3 |
-
"
|
4 |
-
"
|
5 |
-
"train_runtime": 385.1956,
|
6 |
"train_samples": 6750,
|
7 |
-
"train_samples_per_second":
|
8 |
-
"train_steps_per_second": 0.
|
9 |
}
|
|
|
1 |
{
|
2 |
"epoch": 1.0,
|
3 |
+
"train_loss": 0.6385756753525644,
|
4 |
+
"train_runtime": 422.6241,
|
|
|
5 |
"train_samples": 6750,
|
6 |
+
"train_samples_per_second": 15.972,
|
7 |
+
"train_steps_per_second": 0.125
|
8 |
}
|
trainer_state.json
CHANGED
@@ -9,17 +9,13 @@
|
|
9 |
"is_world_process_zero": true,
|
10 |
"log_history": [
|
11 |
{
|
12 |
-
"
|
13 |
-
"debug/policy_weights": 0.49350154399871826,
|
14 |
-
"debug/raw_losses": 0.6931471824645996,
|
15 |
-
"epoch": 0.018867924528301886,
|
16 |
-
"grad_norm": 5.3608531887517,
|
17 |
"learning_rate": 8.333333333333333e-08,
|
18 |
-
"logits/chosen": -2.
|
19 |
-
"logits/rejected": -2.
|
20 |
-
"logps/chosen": -
|
21 |
-
"logps/rejected": -
|
22 |
-
"loss": 0.
|
23 |
"rewards/accuracies": 0.0,
|
24 |
"rewards/chosen": 0.0,
|
25 |
"rewards/margins": 0.0,
|
@@ -27,124 +23,90 @@
|
|
27 |
"step": 1
|
28 |
},
|
29 |
{
|
30 |
-
"
|
31 |
-
"debug/policy_weights": 0.541907548904419,
|
32 |
-
"debug/raw_losses": 0.6909322738647461,
|
33 |
-
"epoch": 0.18867924528301888,
|
34 |
-
"grad_norm": 5.523379066284079,
|
35 |
"learning_rate": 4.911172937635942e-07,
|
36 |
-
"logits/chosen": -2.
|
37 |
-
"logits/rejected": -2.
|
38 |
-
"logps/chosen": -305.
|
39 |
-
"logps/rejected": -
|
40 |
-
"loss": 0.
|
41 |
-
"rewards/accuracies": 0.
|
42 |
-
"rewards/chosen": 0.
|
43 |
-
"rewards/margins": 0.
|
44 |
-
"rewards/rejected": -0.
|
45 |
"step": 10
|
46 |
},
|
47 |
{
|
48 |
-
"
|
49 |
-
"debug/policy_weights": 0.5464785099029541,
|
50 |
-
"debug/raw_losses": 0.6741721630096436,
|
51 |
-
"epoch": 0.37735849056603776,
|
52 |
-
"grad_norm": 5.208106443938916,
|
53 |
"learning_rate": 3.982949361823388e-07,
|
54 |
-
"logits/chosen": -2.
|
55 |
-
"logits/rejected": -2.
|
56 |
-
"logps/chosen": -
|
57 |
-
"logps/rejected": -
|
58 |
-
"loss": 0.
|
59 |
-
"rewards/accuracies": 0.
|
60 |
-
"rewards/chosen": 0.
|
61 |
-
"rewards/margins": 0.
|
62 |
-
"rewards/rejected": -0.
|
63 |
"step": 20
|
64 |
},
|
65 |
{
|
66 |
-
"
|
67 |
-
"debug/policy_weights": 0.563360333442688,
|
68 |
-
"debug/raw_losses": 0.6445623636245728,
|
69 |
-
"epoch": 0.5660377358490566,
|
70 |
-
"grad_norm": 5.583750439900413,
|
71 |
"learning_rate": 2.416462557480814e-07,
|
72 |
-
"logits/chosen": -2.
|
73 |
-
"logits/rejected": -2.
|
74 |
-
"logps/chosen": -
|
75 |
-
"logps/rejected": -
|
76 |
-
"loss": 0.
|
77 |
-
"rewards/accuracies": 0.
|
78 |
-
"rewards/chosen":
|
79 |
-
"rewards/margins": 0.
|
80 |
-
"rewards/rejected": -0.
|
81 |
"step": 30
|
82 |
},
|
83 |
{
|
84 |
-
"
|
85 |
-
"debug/policy_weights": 0.5360943675041199,
|
86 |
-
"debug/raw_losses": 0.5777658224105835,
|
87 |
-
"epoch": 0.7547169811320755,
|
88 |
-
"grad_norm": 4.898347932161407,
|
89 |
"learning_rate": 8.859303711029939e-08,
|
90 |
-
"logits/chosen": -2.
|
91 |
-
"logits/rejected": -2.
|
92 |
-
"logps/chosen": -
|
93 |
-
"logps/rejected": -
|
94 |
-
"loss": 0.
|
95 |
-
"rewards/accuracies": 0.
|
96 |
-
"rewards/chosen":
|
97 |
-
"rewards/margins": 0.
|
98 |
-
"rewards/rejected": -0.
|
99 |
"step": 40
|
100 |
},
|
101 |
{
|
102 |
-
"
|
103 |
-
"debug/policy_weights": 0.5814041495323181,
|
104 |
-
"debug/raw_losses": 0.6445982456207275,
|
105 |
-
"epoch": 0.9433962264150944,
|
106 |
-
"grad_norm": 5.24238516043373,
|
107 |
"learning_rate": 5.009573740853313e-09,
|
108 |
-
"logits/chosen": -2.
|
109 |
-
"logits/rejected": -2.
|
110 |
-
"logps/chosen": -
|
111 |
-
"logps/rejected": -
|
112 |
-
"loss": 0.
|
113 |
-
"rewards/accuracies": 0.
|
114 |
-
"rewards/chosen":
|
115 |
-
"rewards/margins": 0.
|
116 |
-
"rewards/rejected": -0.
|
117 |
"step": 50
|
118 |
},
|
119 |
{
|
120 |
"epoch": 1.0,
|
121 |
"step": 53,
|
122 |
"total_flos": 0.0,
|
123 |
-
"train_loss": 0.
|
124 |
-
"train_runtime":
|
125 |
-
"train_samples_per_second":
|
126 |
-
"train_steps_per_second": 0.
|
127 |
}
|
128 |
],
|
129 |
"logging_steps": 10,
|
130 |
"max_steps": 53,
|
131 |
-
"num_input_tokens_seen": 0,
|
132 |
"num_train_epochs": 1,
|
133 |
"save_steps": 100,
|
134 |
-
"stateful_callbacks": {
|
135 |
-
"TrainerControl": {
|
136 |
-
"args": {
|
137 |
-
"should_epoch_stop": false,
|
138 |
-
"should_evaluate": false,
|
139 |
-
"should_log": false,
|
140 |
-
"should_save": false,
|
141 |
-
"should_training_stop": false
|
142 |
-
},
|
143 |
-
"attributes": {}
|
144 |
-
}
|
145 |
-
},
|
146 |
"total_flos": 0.0,
|
147 |
-
"train_batch_size": 8,
|
148 |
"trial_name": null,
|
149 |
"trial_params": null
|
150 |
}
|
|
|
9 |
"is_world_process_zero": true,
|
10 |
"log_history": [
|
11 |
{
|
12 |
+
"epoch": 0.02,
|
|
|
|
|
|
|
|
|
13 |
"learning_rate": 8.333333333333333e-08,
|
14 |
+
"logits/chosen": -2.8462421894073486,
|
15 |
+
"logits/rejected": -2.8283610343933105,
|
16 |
+
"logps/chosen": -274.7393798828125,
|
17 |
+
"logps/rejected": -204.42575073242188,
|
18 |
+
"loss": 0.6931,
|
19 |
"rewards/accuracies": 0.0,
|
20 |
"rewards/chosen": 0.0,
|
21 |
"rewards/margins": 0.0,
|
|
|
23 |
"step": 1
|
24 |
},
|
25 |
{
|
26 |
+
"epoch": 0.19,
|
|
|
|
|
|
|
|
|
27 |
"learning_rate": 4.911172937635942e-07,
|
28 |
+
"logits/chosen": -2.8527991771698,
|
29 |
+
"logits/rejected": -2.8377315998077393,
|
30 |
+
"logps/chosen": -305.9073181152344,
|
31 |
+
"logps/rejected": -295.8478698730469,
|
32 |
+
"loss": 0.6914,
|
33 |
+
"rewards/accuracies": 0.4513888955116272,
|
34 |
+
"rewards/chosen": 0.0023197412956506014,
|
35 |
+
"rewards/margins": 0.0025084479711949825,
|
36 |
+
"rewards/rejected": -0.00018870655912905931,
|
37 |
"step": 10
|
38 |
},
|
39 |
{
|
40 |
+
"epoch": 0.38,
|
|
|
|
|
|
|
|
|
41 |
"learning_rate": 3.982949361823388e-07,
|
42 |
+
"logits/chosen": -2.859750270843506,
|
43 |
+
"logits/rejected": -2.880180835723877,
|
44 |
+
"logps/chosen": -295.7957458496094,
|
45 |
+
"logps/rejected": -332.6015930175781,
|
46 |
+
"loss": 0.6653,
|
47 |
+
"rewards/accuracies": 0.675000011920929,
|
48 |
+
"rewards/chosen": 0.02228003740310669,
|
49 |
+
"rewards/margins": 0.059415679425001144,
|
50 |
+
"rewards/rejected": -0.037135638296604156,
|
51 |
"step": 20
|
52 |
},
|
53 |
{
|
54 |
+
"epoch": 0.57,
|
|
|
|
|
|
|
|
|
55 |
"learning_rate": 2.416462557480814e-07,
|
56 |
+
"logits/chosen": -2.843632936477661,
|
57 |
+
"logits/rejected": -2.8286781311035156,
|
58 |
+
"logps/chosen": -310.9751892089844,
|
59 |
+
"logps/rejected": -322.77532958984375,
|
60 |
+
"loss": 0.6283,
|
61 |
+
"rewards/accuracies": 0.699999988079071,
|
62 |
+
"rewards/chosen": 0.027804672718048096,
|
63 |
+
"rewards/margins": 0.1940310001373291,
|
64 |
+
"rewards/rejected": -0.1662263423204422,
|
65 |
"step": 30
|
66 |
},
|
67 |
{
|
68 |
+
"epoch": 0.75,
|
|
|
|
|
|
|
|
|
69 |
"learning_rate": 8.859303711029939e-08,
|
70 |
+
"logits/chosen": -2.8175368309020996,
|
71 |
+
"logits/rejected": -2.821326494216919,
|
72 |
+
"logps/chosen": -274.8536682128906,
|
73 |
+
"logps/rejected": -349.11505126953125,
|
74 |
+
"loss": 0.6099,
|
75 |
+
"rewards/accuracies": 0.699999988079071,
|
76 |
+
"rewards/chosen": 0.002673505572602153,
|
77 |
+
"rewards/margins": 0.2130366563796997,
|
78 |
+
"rewards/rejected": -0.21036314964294434,
|
79 |
"step": 40
|
80 |
},
|
81 |
{
|
82 |
+
"epoch": 0.94,
|
|
|
|
|
|
|
|
|
83 |
"learning_rate": 5.009573740853313e-09,
|
84 |
+
"logits/chosen": -2.8560073375701904,
|
85 |
+
"logits/rejected": -2.867896556854248,
|
86 |
+
"logps/chosen": -307.2721862792969,
|
87 |
+
"logps/rejected": -350.257568359375,
|
88 |
+
"loss": 0.5984,
|
89 |
+
"rewards/accuracies": 0.7250000238418579,
|
90 |
+
"rewards/chosen": 0.0023462946992367506,
|
91 |
+
"rewards/margins": 0.33910489082336426,
|
92 |
+
"rewards/rejected": -0.336758553981781,
|
93 |
"step": 50
|
94 |
},
|
95 |
{
|
96 |
"epoch": 1.0,
|
97 |
"step": 53,
|
98 |
"total_flos": 0.0,
|
99 |
+
"train_loss": 0.6385756753525644,
|
100 |
+
"train_runtime": 422.6241,
|
101 |
+
"train_samples_per_second": 15.972,
|
102 |
+
"train_steps_per_second": 0.125
|
103 |
}
|
104 |
],
|
105 |
"logging_steps": 10,
|
106 |
"max_steps": 53,
|
|
|
107 |
"num_train_epochs": 1,
|
108 |
"save_steps": 100,
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
109 |
"total_flos": 0.0,
|
|
|
110 |
"trial_name": null,
|
111 |
"trial_params": null
|
112 |
}
|
training_args.bin
CHANGED
@@ -1,3 +1,3 @@
|
|
1 |
version https://git-lfs.github.com/spec/v1
|
2 |
-
oid sha256:
|
3 |
-
size
|
|
|
1 |
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:ebd548d73bc0ddb09d45f8d7707ec72370839df1501f0c2de2d460165f54d13e
|
3 |
+
size 5944
|