File size: 1,686 Bytes
76cebef
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
tokenizer:
  _component_: torchtune.models.mistral.mistral_tokenizer
  path: /tmp/Mistral-7B-Instruct-v0.2/tokenizer.model
dataset:
  _component_: torchtune.datasets.instruct_dataset
  source: prithviraj-maurya/legalbench-entire
  template: AlpacaInstructTemplate
  column_map:
    instruction: instruction
    input: question
    output: answer
  max_seq_len: 256
  train_on_input: true
  split: train
seed: null
shuffle: true
model:
  _component_: torchtune.models.mistral.qlora_mistral_7b
  lora_attn_modules:
  - q_proj
  - k_proj
  - v_proj
  apply_lora_to_mlp: true
  apply_lora_to_output: false
  lora_rank: 64
  lora_alpha: 16
checkpointer:
  _component_: torchtune.utils.FullModelHFCheckpointer
  checkpoint_dir: /tmp/Mistral-7B-Instruct-v0.2
  checkpoint_files:
  - pytorch_model-00001-of-00003.bin
  - pytorch_model-00002-of-00003.bin
  - pytorch_model-00003-of-00003.bin
  recipe_checkpoint: null
  output_dir: /tmp/Mistral-7B-Instruct-v0.2
  model_type: MISTRAL
resume_from_checkpoint: false
optimizer:
  _component_: torch.optim.AdamW
  lr: 2.0e-05
lr_scheduler:
  _component_: torchtune.modules.get_cosine_schedule_with_warmup
  num_warmup_steps: 100
loss:
  _component_: torch.nn.CrossEntropyLoss
batch_size: 4
epochs: 10
max_steps_per_epoch: 100000
gradient_accumulation_steps: 4
compile: false
device: cuda
enable_activation_checkpointing: true
dtype: fp32
output_dir: /logs/mistral_7b_qlora_single_device_finetune
metric_logger:
  _component_: torchtune.utils.metric_logging.WandBLogger
  project: torchtune
log_every_n_steps: 10
profiler:
  _component_: torchtune.utils.profiler
  enabled: false
  output_dir: /tmp/alpaca-llama2-finetune/torchtune_perf_tracing.json