helixx999 commited on
Commit
8783b54
1 Parent(s): faef20e

Update README.md

Browse files
Files changed (1) hide show
  1. README.md +3 -12
README.md CHANGED
@@ -19,19 +19,11 @@ tags:
19
 
20
  This is gemma2 trained on semeval restaurant data 2014 using unsloth framework.
21
 
22
- `trainer = SFTTrainer(
23
- model = model,
24
- tokenizer = tokenizer,
25
- train_dataset = dataset,
26
- dataset_text_field = "text_new",
27
- max_seq_length = max_seq_length,
28
- dataset_num_proc = 2,
29
- packing = False, # Can make training 5x faster for short sequences.
30
- args = TrainingArguments(
31
  per_device_train_batch_size = 2,
32
  gradient_accumulation_steps = 4,
33
  warmup_steps = 6, #Previous 5
34
- #num_train_epochs = 1, # Set this for 1 full training run.
35
  max_steps = 60,
36
  #learning_rate = 2e-4,
37
  learning_rate = 1e-4,
@@ -44,7 +36,6 @@ This is gemma2 trained on semeval restaurant data 2014 using unsloth framework.
44
  seed = 3407,
45
  output_dir = "./tensorLog",
46
  report_to="wandb"
47
- ),
48
- )`
49
 
50
  [<img src="https://raw.githubusercontent.com/unslothai/unsloth/main/images/unsloth%20made%20with%20love.png" width="200"/>](https://github.com/unslothai/unsloth)
 
19
 
20
  This is gemma2 trained on semeval restaurant data 2014 using unsloth framework.
21
 
22
+ Training Parameters:
23
+
 
 
 
 
 
 
 
24
  per_device_train_batch_size = 2,
25
  gradient_accumulation_steps = 4,
26
  warmup_steps = 6, #Previous 5
 
27
  max_steps = 60,
28
  #learning_rate = 2e-4,
29
  learning_rate = 1e-4,
 
36
  seed = 3407,
37
  output_dir = "./tensorLog",
38
  report_to="wandb"
39
+
 
40
 
41
  [<img src="https://raw.githubusercontent.com/unslothai/unsloth/main/images/unsloth%20made%20with%20love.png" width="200"/>](https://github.com/unslothai/unsloth)