tuanna08go commited on
Commit
a6b0546
·
verified ·
1 Parent(s): a144718

End of training

Browse files
Files changed (2) hide show
  1. README.md +11 -4
  2. adapter_model.bin +1 -1
README.md CHANGED
@@ -66,7 +66,7 @@ lora_model_dir: null
66
  lora_r: 8
67
  lora_target_linear: true
68
  lr_scheduler: cosine
69
- max_steps: 1
70
  micro_batch_size: 2
71
  mlflow_experiment_name: /tmp/ab14738766939fdd_train_data.json
72
  model_type: AutoModelForCausalLM
@@ -93,7 +93,7 @@ wandb_name: 59f579bd-744f-4da4-b1bd-9709a58ffa93
93
  wandb_project: Gradients-On-Demand
94
  wandb_run: your_name
95
  wandb_runid: 59f579bd-744f-4da4-b1bd-9709a58ffa93
96
- warmup_steps: 1
97
  weight_decay: 0.0
98
  xformers_attention: null
99
 
@@ -104,6 +104,8 @@ xformers_attention: null
104
  # 89ee1f7a-343a-4eb6-883d-79428bf1b372
105
 
106
  This model is a fine-tuned version of [EleutherAI/pythia-160m](https://huggingface.co/EleutherAI/pythia-160m) on the None dataset.
 
 
107
 
108
  ## Model description
109
 
@@ -130,14 +132,19 @@ The following hyperparameters were used during training:
130
  - total_train_batch_size: 8
131
  - optimizer: Use OptimizerNames.ADAMW_BNB with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
132
  - lr_scheduler_type: cosine
133
- - lr_scheduler_warmup_steps: 2
134
- - training_steps: 1
135
 
136
  ### Training results
137
 
138
  | Training Loss | Epoch | Step | Validation Loss |
139
  |:-------------:|:------:|:----:|:---------------:|
140
  | No log | 0.0001 | 1 | 0.9816 |
 
 
 
 
 
141
 
142
 
143
  ### Framework versions
 
66
  lora_r: 8
67
  lora_target_linear: true
68
  lr_scheduler: cosine
69
+ max_steps: 50
70
  micro_batch_size: 2
71
  mlflow_experiment_name: /tmp/ab14738766939fdd_train_data.json
72
  model_type: AutoModelForCausalLM
 
93
  wandb_project: Gradients-On-Demand
94
  wandb_run: your_name
95
  wandb_runid: 59f579bd-744f-4da4-b1bd-9709a58ffa93
96
+ warmup_steps: 10
97
  weight_decay: 0.0
98
  xformers_attention: null
99
 
 
104
  # 89ee1f7a-343a-4eb6-883d-79428bf1b372
105
 
106
  This model is a fine-tuned version of [EleutherAI/pythia-160m](https://huggingface.co/EleutherAI/pythia-160m) on the None dataset.
107
+ It achieves the following results on the evaluation set:
108
+ - Loss: 0.7336
109
 
110
  ## Model description
111
 
 
132
  - total_train_batch_size: 8
133
  - optimizer: Use OptimizerNames.ADAMW_BNB with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
134
  - lr_scheduler_type: cosine
135
+ - lr_scheduler_warmup_steps: 10
136
+ - training_steps: 50
137
 
138
  ### Training results
139
 
140
  | Training Loss | Epoch | Step | Validation Loss |
141
  |:-------------:|:------:|:----:|:---------------:|
142
  | No log | 0.0001 | 1 | 0.9816 |
143
+ | 3.8414 | 0.0011 | 10 | 0.9614 |
144
+ | 3.5844 | 0.0022 | 20 | 0.8447 |
145
+ | 2.878 | 0.0034 | 30 | 0.7686 |
146
+ | 3.0867 | 0.0045 | 40 | 0.7385 |
147
+ | 2.8059 | 0.0056 | 50 | 0.7336 |
148
 
149
 
150
  ### Framework versions
adapter_model.bin CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:0adc944b4d1984ea8e16b45ee4e66406d92b63a9f775f76791b4d3a50a369772
3
  size 4753742
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:94cd99dcdebbad26ac3e1f16572537d9f575ab41770a8e4697d2b4003ca7c74e
3
  size 4753742