Hugging Face's logo

izzcw
/
llama_3_70b_lora_sft_construction

PEFT
Safetensors
llama-factory
lora
Generated from Trainer
Model card Files
xet
 Community
izzcw commited on
Commit
889044f
·
verified ·
1 Parent(s): 3b9c782

End of training

Browse files
Files changed (7) hide show
  1. README.md +3 -2
  2. all_results.json +12 -0
  3. eval_results.json +7 -0
  4. train_results.json +8 -0
  5. trainer_state.json +107 -0
  6. training_eval_loss.png +0 -0
  7. training_loss.png +0 -0
README.md CHANGED
@@ -4,6 +4,7 @@ license: llama3
4
  base_model: meta-llama/Meta-Llama-3-70B-Instruct
5
  tags:
6
  - llama-factory
 
7
  - generated_from_trainer
8
  model-index:
9
  - name: llama_3_70b_lora_sft_construction
@@ -15,9 +16,9 @@ should probably proofread and complete it, then remove this comment. -->
15
 
16
  # llama_3_70b_lora_sft_construction
17
 
18
- This model is a fine-tuned version of [meta-llama/Meta-Llama-3-70B-Instruct](https://huggingface.co/meta-llama/Meta-Llama-3-70B-Instruct) on the None dataset.
19
  It achieves the following results on the evaluation set:
20
- - Loss: 0.3483
21
 
22
  ## Model description
23
 
 
4
  base_model: meta-llama/Meta-Llama-3-70B-Instruct
5
  tags:
6
  - llama-factory
7
+ - lora
8
  - generated_from_trainer
9
  model-index:
10
  - name: llama_3_70b_lora_sft_construction
 
16
 
17
  # llama_3_70b_lora_sft_construction
18
 
19
+ This model is a fine-tuned version of [meta-llama/Meta-Llama-3-70B-Instruct](https://huggingface.co/meta-llama/Meta-Llama-3-70B-Instruct) on the identity and the filtered_construction_train_data datasets.
20
  It achieves the following results on the evaluation set:
21
+ - Loss: 0.3475
22
 
23
  ## Model description
24
 
all_results.json ADDED
@@ -0,0 +1,12 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "epoch": 0.9947460595446584,
3
+ "eval_loss": 0.34747257828712463,
4
+ "eval_runtime": 38.5498,
5
+ "eval_samples_per_second": 4.851,
6
+ "eval_steps_per_second": 0.623,
7
+ "total_flos": 8671839626723328.0,
8
+ "train_loss": 0.3312246486754485,
9
+ "train_runtime": 6300.0732,
10
+ "train_samples_per_second": 1.45,
11
+ "train_steps_per_second": 0.011
12
+ }
eval_results.json ADDED
@@ -0,0 +1,7 @@
 
 
 
 
 
 
 
 
1
+ {
2
+ "epoch": 0.9947460595446584,
3
+ "eval_loss": 0.34747257828712463,
4
+ "eval_runtime": 38.5498,
5
+ "eval_samples_per_second": 4.851,
6
+ "eval_steps_per_second": 0.623
7
+ }
train_results.json ADDED
@@ -0,0 +1,8 @@
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "epoch": 0.9947460595446584,
3
+ "total_flos": 8671839626723328.0,
4
+ "train_loss": 0.3312246486754485,
5
+ "train_runtime": 6300.0732,
6
+ "train_samples_per_second": 1.45,
7
+ "train_steps_per_second": 0.011
8
+ }
trainer_state.json ADDED
@@ -0,0 +1,107 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "best_metric": null,
3
+ "best_model_checkpoint": null,
4
+ "epoch": 0.9947460595446584,
5
+ "eval_steps": 30,
6
+ "global_step": 71,
7
+ "is_hyper_param_search": false,
8
+ "is_local_process_zero": true,
9
+ "is_world_process_zero": true,
10
+ "log_history": [
11
+ {
12
+ "epoch": 0.14010507880910683,
13
+ "grad_norm": 0.45420220075243706,
14
+ "learning_rate": 9.975153876827008e-05,
15
+ "loss": 0.5917,
16
+ "step": 10
17
+ },
18
+ {
19
+ "epoch": 0.28021015761821366,
20
+ "grad_norm": 0.4281694659811335,
21
+ "learning_rate": 9.131193871579975e-05,
22
+ "loss": 0.3715,
23
+ "step": 20
24
+ },
25
+ {
26
+ "epoch": 0.4203152364273205,
27
+ "grad_norm": 0.2752285164412833,
28
+ "learning_rate": 7.281053286765815e-05,
29
+ "loss": 0.2844,
30
+ "step": 30
31
+ },
32
+ {
33
+ "epoch": 0.4203152364273205,
34
+ "eval_loss": 0.3744376003742218,
35
+ "eval_runtime": 38.8917,
36
+ "eval_samples_per_second": 4.808,
37
+ "eval_steps_per_second": 0.617,
38
+ "step": 30
39
+ },
40
+ {
41
+ "epoch": 0.5604203152364273,
42
+ "grad_norm": 0.2616196043777537,
43
+ "learning_rate": 4.875346541309637e-05,
44
+ "loss": 0.2776,
45
+ "step": 40
46
+ },
47
+ {
48
+ "epoch": 0.7005253940455342,
49
+ "grad_norm": 0.2979118016206158,
50
+ "learning_rate": 2.500000000000001e-05,
51
+ "loss": 0.2932,
52
+ "step": 50
53
+ },
54
+ {
55
+ "epoch": 0.840630472854641,
56
+ "grad_norm": 0.2227457988487204,
57
+ "learning_rate": 7.33545591839222e-06,
58
+ "loss": 0.2646,
59
+ "step": 60
60
+ },
61
+ {
62
+ "epoch": 0.840630472854641,
63
+ "eval_loss": 0.3483392000198364,
64
+ "eval_runtime": 38.9637,
65
+ "eval_samples_per_second": 4.799,
66
+ "eval_steps_per_second": 0.616,
67
+ "step": 60
68
+ },
69
+ {
70
+ "epoch": 0.9807355516637478,
71
+ "grad_norm": 0.39490090028744906,
72
+ "learning_rate": 6.215393905388278e-08,
73
+ "loss": 0.2557,
74
+ "step": 70
75
+ },
76
+ {
77
+ "epoch": 0.9947460595446584,
78
+ "step": 71,
79
+ "total_flos": 8671839626723328.0,
80
+ "train_loss": 0.3312246486754485,
81
+ "train_runtime": 6300.0732,
82
+ "train_samples_per_second": 1.45,
83
+ "train_steps_per_second": 0.011
84
+ }
85
+ ],
86
+ "logging_steps": 10,
87
+ "max_steps": 71,
88
+ "num_input_tokens_seen": 0,
89
+ "num_train_epochs": 1,
90
+ "save_steps": 500,
91
+ "stateful_callbacks": {
92
+ "TrainerControl": {
93
+ "args": {
94
+ "should_epoch_stop": false,
95
+ "should_evaluate": false,
96
+ "should_log": false,
97
+ "should_save": true,
98
+ "should_training_stop": true
99
+ },
100
+ "attributes": {}
101
+ }
102
+ },
103
+ "total_flos": 8671839626723328.0,
104
+ "train_batch_size": 1,
105
+ "trial_name": null,
106
+ "trial_params": null
107
+ }
training_eval_loss.png ADDED
training_loss.png ADDED
Free AI Image Generator No sign-up. Instant results. Open Now