Hugging Face's logo

ericrisco
/
llama2_instruct_generation

PEFT
TensorBoard
Safetensors
llama
trl
sft
Generated from Trainer
4-bit precision
bitsandbytes
Model card Files Metrics Community
ericrisco commited on
Commit
284c81c
·
verified ·
1 Parent(s): ffb8dd6

ericrisco/llama2-instruct-tune-500s

Browse files
Files changed (4) hide show
  1. README.md +23 -23
  2. adapter_model.safetensors +1 -1
  3. runs/Sep26_13-42-54_3d63c6bcbcfd/events.out.tfevents.1727358175.3d63c6bcbcfd.3916.1 +3 -0
  4. training_args.bin +1 -1
README.md CHANGED
@@ -19,7 +19,7 @@ should probably proofread and complete it, then remove this comment. -->
19
 
20
  This model is a fine-tuned version of [NousResearch/Llama-2-7b-hf](https://huggingface.co/NousResearch/Llama-2-7b-hf) on the generator dataset.
21
  It achieves the following results on the evaluation set:
22
- - Loss: 1.6761
23
 
24
  ## Model description
25
 
@@ -51,30 +51,30 @@ The following hyperparameters were used during training:
51
  | Training Loss | Epoch | Step | Validation Loss |
52
  |:-------------:|:------:|:----:|:---------------:|
53
  | 1.9468 | 0.0027 | 20 | 1.8141 |
54
- | 1.8737 | 0.0054 | 40 | 1.7848 |
55
  | 1.8769 | 0.0081 | 60 | 1.7718 |
56
- | 1.8634 | 0.0108 | 80 | 1.7598 |
57
- | 1.8584 | 0.0135 | 100 | 1.7469 |
58
- | 1.8271 | 0.0163 | 120 | 1.7170 |
59
- | 1.8706 | 0.0190 | 140 | 1.7042 |
60
- | 1.8306 | 0.0217 | 160 | 1.7005 |
61
- | 1.7954 | 0.0244 | 180 | 1.6948 |
62
- | 1.8616 | 0.0271 | 200 | 1.6947 |
63
  | 1.81 | 0.0298 | 220 | 1.6915 |
64
- | 1.8003 | 0.0325 | 240 | 1.6900 |
65
- | 1.9069 | 0.0352 | 260 | 1.6880 |
66
- | 1.8266 | 0.0379 | 280 | 1.6868 |
67
- | 1.8615 | 0.0406 | 300 | 1.6849 |
68
- | 1.7728 | 0.0433 | 320 | 1.6832 |
69
- | 1.806 | 0.0461 | 340 | 1.6824 |
70
- | 1.8843 | 0.0488 | 360 | 1.6812 |
71
- | 1.7655 | 0.0515 | 380 | 1.6803 |
72
- | 1.812 | 0.0542 | 400 | 1.6795 |
73
- | 1.8058 | 0.0569 | 420 | 1.6779 |
74
- | 1.7424 | 0.0596 | 440 | 1.6779 |
75
- | 1.8976 | 0.0623 | 460 | 1.6782 |
76
- | 1.8237 | 0.0650 | 480 | 1.6778 |
77
- | 1.8981 | 0.0677 | 500 | 1.6761 |
78
 
79
 
80
  ### Framework versions
 
19
 
20
  This model is a fine-tuned version of [NousResearch/Llama-2-7b-hf](https://huggingface.co/NousResearch/Llama-2-7b-hf) on the generator dataset.
21
  It achieves the following results on the evaluation set:
22
+ - Loss: 1.6759
23
 
24
  ## Model description
25
 
 
51
  | Training Loss | Epoch | Step | Validation Loss |
52
  |:-------------:|:------:|:----:|:---------------:|
53
  | 1.9468 | 0.0027 | 20 | 1.8141 |
54
+ | 1.8734 | 0.0054 | 40 | 1.7849 |
55
  | 1.8769 | 0.0081 | 60 | 1.7718 |
56
+ | 1.8633 | 0.0108 | 80 | 1.7599 |
57
+ | 1.8583 | 0.0135 | 100 | 1.7472 |
58
+ | 1.8264 | 0.0163 | 120 | 1.7176 |
59
+ | 1.8714 | 0.0190 | 140 | 1.7053 |
60
+ | 1.831 | 0.0217 | 160 | 1.7012 |
61
+ | 1.7957 | 0.0244 | 180 | 1.6947 |
62
+ | 1.8613 | 0.0271 | 200 | 1.6934 |
63
  | 1.81 | 0.0298 | 220 | 1.6915 |
64
+ | 1.7995 | 0.0325 | 240 | 1.6893 |
65
+ | 1.9067 | 0.0352 | 260 | 1.6872 |
66
+ | 1.8261 | 0.0379 | 280 | 1.6860 |
67
+ | 1.8609 | 0.0406 | 300 | 1.6843 |
68
+ | 1.7725 | 0.0433 | 320 | 1.6835 |
69
+ | 1.8061 | 0.0461 | 340 | 1.6819 |
70
+ | 1.8842 | 0.0488 | 360 | 1.6804 |
71
+ | 1.7648 | 0.0515 | 380 | 1.6799 |
72
+ | 1.8121 | 0.0542 | 400 | 1.6796 |
73
+ | 1.8056 | 0.0569 | 420 | 1.6777 |
74
+ | 1.7423 | 0.0596 | 440 | 1.6780 |
75
+ | 1.8971 | 0.0623 | 460 | 1.6782 |
76
+ | 1.8234 | 0.0650 | 480 | 1.6771 |
77
+ | 1.8978 | 0.0677 | 500 | 1.6759 |
78
 
79
 
80
  ### Framework versions
adapter_model.safetensors CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:c43dc04532b014d1c7de792f57ae7b7d5a3f8e6bd497a9677993c63d5ee08096
3
  size 134235048
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:311ca013ee3478a8b5594c05f6312e32c1a788788fbfac207b6914e873bea033
3
  size 134235048
runs/Sep26_13-42-54_3d63c6bcbcfd/events.out.tfevents.1727358175.3d63c6bcbcfd.3916.1 ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:2f8aa56e0ed8aeb8b9fe65753f25dc6673bdbfef547196ed138c9db603ad9277
3
+ size 23452
training_args.bin CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:0d7d05ef4184ef5384d4cbcc938ea6aefb352a1c32c3276f279e3f32f1fe21f7
3
  size 5496
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:837ee62862faf7e261a706baf6d07f66d3d7cadc6beb5e43fc4a5fc76ec3d4c4
3
  size 5496