Jerry46 commited on
Commit
ce8dc73
·
1 Parent(s): 5d1b55b

Training in progress, epoch 0

Browse files
README.md CHANGED
@@ -15,15 +15,15 @@ should probably proofread and complete it, then remove this comment. -->
15
 
16
  This model is a fine-tuned version of [mistralai/Mistral-7B-v0.1](https://huggingface.co/mistralai/Mistral-7B-v0.1) on the None dataset.
17
  It achieves the following results on the evaluation set:
18
- - Loss: 0.6642
19
- - Rewards/chosen: 0.1042
20
- - Rewards/rejected: 0.0401
21
- - Rewards/accuracies: 0.6480
22
- - Rewards/margins: 0.0641
23
- - Logps/rejected: -230.4560
24
- - Logps/chosen: -278.6917
25
- - Logits/rejected: -2.3987
26
- - Logits/chosen: -2.4597
27
 
28
  ## Model description
29
 
@@ -43,24 +43,26 @@ More information needed
43
 
44
  The following hyperparameters were used during training:
45
  - learning_rate: 5e-07
46
- - train_batch_size: 8
47
  - eval_batch_size: 4
48
  - seed: 42
49
  - distributed_type: multi-GPU
50
  - num_devices: 4
51
- - gradient_accumulation_steps: 2
52
  - total_train_batch_size: 64
53
  - total_eval_batch_size: 16
54
  - optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
55
  - lr_scheduler_type: linear
56
  - lr_scheduler_warmup_ratio: 0.1
57
- - num_epochs: 1
58
 
59
  ### Training results
60
 
61
  | Training Loss | Epoch | Step | Validation Loss | Rewards/chosen | Rewards/rejected | Rewards/accuracies | Rewards/margins | Logps/rejected | Logps/chosen | Logits/rejected | Logits/chosen |
62
  |:-------------:|:-----:|:----:|:---------------:|:--------------:|:----------------:|:------------------:|:---------------:|:--------------:|:------------:|:---------------:|:-------------:|
63
- | 0.661 | 1.0 | 968 | 0.6642 | 0.1042 | 0.0401 | 0.6480 | 0.0641 | -230.4560 | -278.6917 | -2.3987 | -2.4597 |
 
 
64
 
65
 
66
  ### Framework versions
 
15
 
16
  This model is a fine-tuned version of [mistralai/Mistral-7B-v0.1](https://huggingface.co/mistralai/Mistral-7B-v0.1) on the None dataset.
17
  It achieves the following results on the evaluation set:
18
+ - Loss: 0.5266
19
+ - Rewards/chosen: -0.1470
20
+ - Rewards/rejected: -0.8985
21
+ - Rewards/accuracies: 0.7460
22
+ - Rewards/margins: 0.7516
23
+ - Logps/rejected: -228.2694
24
+ - Logps/chosen: -266.1304
25
+ - Logits/rejected: -1.9412
26
+ - Logits/chosen: -2.0659
27
 
28
  ## Model description
29
 
 
43
 
44
  The following hyperparameters were used during training:
45
  - learning_rate: 5e-07
46
+ - train_batch_size: 2
47
  - eval_batch_size: 4
48
  - seed: 42
49
  - distributed_type: multi-GPU
50
  - num_devices: 4
51
+ - gradient_accumulation_steps: 8
52
  - total_train_batch_size: 64
53
  - total_eval_batch_size: 16
54
  - optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
55
  - lr_scheduler_type: linear
56
  - lr_scheduler_warmup_ratio: 0.1
57
+ - num_epochs: 3
58
 
59
  ### Training results
60
 
61
  | Training Loss | Epoch | Step | Validation Loss | Rewards/chosen | Rewards/rejected | Rewards/accuracies | Rewards/margins | Logps/rejected | Logps/chosen | Logits/rejected | Logits/chosen |
62
  |:-------------:|:-----:|:----:|:---------------:|:--------------:|:----------------:|:------------------:|:---------------:|:--------------:|:------------:|:---------------:|:-------------:|
63
+ | 0.5501 | 1.0 | 968 | 0.5560 | -0.1093 | -0.6400 | 0.7200 | 0.5307 | -225.6847 | -265.7538 | -1.9901 | -2.1115 |
64
+ | 0.5412 | 2.0 | 1936 | 0.5318 | -0.1497 | -0.8640 | 0.7420 | 0.7143 | -227.9245 | -266.1583 | -1.9509 | -2.0748 |
65
+ | 0.5454 | 3.0 | 2904 | 0.5266 | -0.1470 | -0.8985 | 0.7460 | 0.7516 | -228.2694 | -266.1304 | -1.9412 | -2.0659 |
66
 
67
 
68
  ### Framework versions
adapter_config.json CHANGED
@@ -16,9 +16,9 @@
16
  "rank_pattern": {},
17
  "revision": null,
18
  "target_modules": [
 
19
  "v_proj",
20
  "q_proj",
21
- "o_proj",
22
  "k_proj"
23
  ],
24
  "task_type": "CAUSAL_LM"
 
16
  "rank_pattern": {},
17
  "revision": null,
18
  "target_modules": [
19
+ "o_proj",
20
  "v_proj",
21
  "q_proj",
 
22
  "k_proj"
23
  ],
24
  "task_type": "CAUSAL_LM"
adapter_model.safetensors CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:e37e1bc399e9cc116250e3e649a446d8bc7c7fc9e918b00f398d5cf4eb9b9a71
3
- size 109086672
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:9c393ebd4252a03b80a5249d58bc2098224b1be8de7d9bef2676336f9dd92052
3
+ size 218138576
all_results.json CHANGED
@@ -1,21 +1,21 @@
1
  {
2
- "epoch": 1.0,
3
- "eval_logits/chosen": -2.4597132205963135,
4
- "eval_logits/rejected": -2.398695468902588,
5
- "eval_logps/chosen": -278.69171142578125,
6
- "eval_logps/rejected": -230.4560089111328,
7
- "eval_loss": 0.6642152070999146,
8
- "eval_rewards/accuracies": 0.6480000019073486,
9
- "eval_rewards/chosen": 0.10415761172771454,
10
- "eval_rewards/margins": 0.06405296921730042,
11
- "eval_rewards/rejected": 0.04010463133454323,
12
- "eval_runtime": 444.8959,
13
  "eval_samples": 2000,
14
- "eval_samples_per_second": 4.495,
15
- "eval_steps_per_second": 0.281,
16
- "train_loss": 0.6728762634529555,
17
- "train_runtime": 27528.1814,
18
  "train_samples": 61966,
19
- "train_samples_per_second": 2.251,
20
- "train_steps_per_second": 0.035
21
  }
 
1
  {
2
+ "epoch": 3.0,
3
+ "eval_logits/chosen": -2.0658671855926514,
4
+ "eval_logits/rejected": -1.9411793947219849,
5
+ "eval_logps/chosen": -266.13037109375,
6
+ "eval_logps/rejected": -228.2694091796875,
7
+ "eval_loss": 0.5265706777572632,
8
+ "eval_rewards/accuracies": 0.7459999918937683,
9
+ "eval_rewards/chosen": -0.1469534933567047,
10
+ "eval_rewards/margins": 0.7515553832054138,
11
+ "eval_rewards/rejected": -0.8985088467597961,
12
+ "eval_runtime": 453.4186,
13
  "eval_samples": 2000,
14
+ "eval_samples_per_second": 4.411,
15
+ "eval_steps_per_second": 0.276,
16
+ "train_loss": 0.5642705829018732,
17
+ "train_runtime": 71197.0849,
18
  "train_samples": 61966,
19
+ "train_samples_per_second": 2.611,
20
+ "train_steps_per_second": 0.041
21
  }
eval_results.json CHANGED
@@ -1,16 +1,16 @@
1
  {
2
- "epoch": 1.0,
3
- "eval_logits/chosen": -2.4597132205963135,
4
- "eval_logits/rejected": -2.398695468902588,
5
- "eval_logps/chosen": -278.69171142578125,
6
- "eval_logps/rejected": -230.4560089111328,
7
- "eval_loss": 0.6642152070999146,
8
- "eval_rewards/accuracies": 0.6480000019073486,
9
- "eval_rewards/chosen": 0.10415761172771454,
10
- "eval_rewards/margins": 0.06405296921730042,
11
- "eval_rewards/rejected": 0.04010463133454323,
12
- "eval_runtime": 444.8959,
13
  "eval_samples": 2000,
14
- "eval_samples_per_second": 4.495,
15
- "eval_steps_per_second": 0.281
16
  }
 
1
  {
2
+ "epoch": 3.0,
3
+ "eval_logits/chosen": -2.0658671855926514,
4
+ "eval_logits/rejected": -1.9411793947219849,
5
+ "eval_logps/chosen": -266.13037109375,
6
+ "eval_logps/rejected": -228.2694091796875,
7
+ "eval_loss": 0.5265706777572632,
8
+ "eval_rewards/accuracies": 0.7459999918937683,
9
+ "eval_rewards/chosen": -0.1469534933567047,
10
+ "eval_rewards/margins": 0.7515553832054138,
11
+ "eval_rewards/rejected": -0.8985088467597961,
12
+ "eval_runtime": 453.4186,
13
  "eval_samples": 2000,
14
+ "eval_samples_per_second": 4.411,
15
+ "eval_steps_per_second": 0.276
16
  }
runs/Dec09_14-07-41_uclaml03.cs.ucla.edu/events.out.tfevents.1702159748.uclaml03.cs.ucla.edu.2757530.0 ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:6cea1ae3de70d8451e6f46b3c4461dfcb3d37990939d1372ff2b9ddb87567759
3
+ size 4958
runs/Dec09_15-37-03_uclaml03.cs.ucla.edu/events.out.tfevents.1702165262.uclaml03.cs.ucla.edu.2770644.0 ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:e45658132378c248980f03ceb5004b8fb63991772648afca77ac7acbbefc4eae
3
+ size 4971
runs/Dec09_15-53-30_uclaml03.cs.ucla.edu/events.out.tfevents.1702166063.uclaml03.cs.ucla.edu.2772424.0 ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:faa4263bf9f2ecaca5198fc1674069d502889c76c31205a5acb8bebd0b4c2844
3
+ size 4348
runs/Dec09_15-56-00_uclaml03.cs.ucla.edu/events.out.tfevents.1702166213.uclaml03.cs.ucla.edu.2772782.0 ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:c6e64ebc53833e9e1e25dcb8c992edca6578e44e31cbfca2464515fe61b8f3c9
3
+ size 4970
runs/Dec09_15-58-05_uclaml03.cs.ucla.edu/events.out.tfevents.1702166345.uclaml03.cs.ucla.edu.2773187.0 ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:f2ea0fa0304b57f9ddf922d8b6095cdf101e2504972fa84b96d0f97aef5cc4a8
3
+ size 8708
runs/Dec13_15-13-52_uclaml03.cs.ucla.edu/events.out.tfevents.1702509291.uclaml03.cs.ucla.edu.3092760.0 ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:7e67b7e41177363fc824fe7bbeaa95ad84a54895c764d5c23511b2045127770a
3
+ size 5622
runs/Dec13_15-25-44_uclaml03.cs.ucla.edu/events.out.tfevents.1702510148.uclaml03.cs.ucla.edu.3094406.0 ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:776803458c24dc1fbf0b0b727d4dd63f7ce254f862bebc801e99a54f14d5e845
3
+ size 20786
train_results.json CHANGED
@@ -1,8 +1,8 @@
1
  {
2
- "epoch": 1.0,
3
- "train_loss": 0.6728762634529555,
4
- "train_runtime": 27528.1814,
5
  "train_samples": 61966,
6
- "train_samples_per_second": 2.251,
7
- "train_steps_per_second": 0.035
8
  }
 
1
  {
2
+ "epoch": 3.0,
3
+ "train_loss": 0.5642705829018732,
4
+ "train_runtime": 71197.0849,
5
  "train_samples": 61966,
6
+ "train_samples_per_second": 2.611,
7
+ "train_steps_per_second": 0.041
8
  }
trainer_state.json CHANGED
The diff for this file is too large to render. See raw diff
 
training_args.bin CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:7ad26a1d13fc0eb39a1c99b55d10b9142a6919aa345c52de57a8a403cdd0c317
3
- size 4792
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:39f0543027e85b9a836fd74541ae3c0b6e2c0ddbebf81a439c49f97cd7639bd4
3
+ size 4728