Commit
•
17ba267
1
Parent(s):
c32fd37
Model save
Browse files- README.md +3 -27
- adapter_model.safetensors +1 -1
README.md
CHANGED
|
@@ -6,8 +6,6 @@ tags:
|
|
| 6 |
- sft
|
| 7 |
- generated_from_trainer
|
| 8 |
base_model: google/gemma-2b
|
| 9 |
-
datasets:
|
| 10 |
-
- generator
|
| 11 |
model-index:
|
| 12 |
- name: Not-so-bright-AGI-v1
|
| 13 |
results: []
|
|
@@ -18,9 +16,7 @@ should probably proofread and complete it, then remove this comment. -->
|
|
| 18 |
|
| 19 |
# Not-so-bright-AGI-v1
|
| 20 |
|
| 21 |
-
This model is a fine-tuned version of [google/gemma-2b](https://huggingface.co/google/gemma-2b) on
|
| 22 |
-
It achieves the following results on the evaluation set:
|
| 23 |
-
- Loss: 2.0209
|
| 24 |
|
| 25 |
## Model description
|
| 26 |
|
|
@@ -40,36 +36,16 @@ More information needed
|
|
| 40 |
|
| 41 |
The following hyperparameters were used during training:
|
| 42 |
- learning_rate: 1e-05
|
| 43 |
-
- train_batch_size:
|
| 44 |
- eval_batch_size: 8
|
| 45 |
- seed: 42
|
| 46 |
- gradient_accumulation_steps: 8
|
| 47 |
-
- total_train_batch_size:
|
| 48 |
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
|
| 49 |
- lr_scheduler_type: linear
|
| 50 |
- lr_scheduler_warmup_ratio: 0.05
|
| 51 |
- training_steps: 1480
|
| 52 |
|
| 53 |
-
### Training results
|
| 54 |
-
|
| 55 |
-
| Training Loss | Epoch | Step | Validation Loss |
|
| 56 |
-
|:-------------:|:-------:|:----:|:---------------:|
|
| 57 |
-
| 2.9261 | 1.6393 | 100 | 2.5691 |
|
| 58 |
-
| 2.4363 | 3.2787 | 200 | 2.2789 |
|
| 59 |
-
| 2.2448 | 4.9180 | 300 | 2.1604 |
|
| 60 |
-
| 2.1502 | 6.5574 | 400 | 2.1008 |
|
| 61 |
-
| 2.111 | 8.1967 | 500 | 2.0725 |
|
| 62 |
-
| 2.0837 | 9.8361 | 600 | 2.0565 |
|
| 63 |
-
| 2.0646 | 11.4754 | 700 | 2.0456 |
|
| 64 |
-
| 2.0499 | 13.1148 | 800 | 2.0378 |
|
| 65 |
-
| 2.0497 | 14.7541 | 900 | 2.0323 |
|
| 66 |
-
| 2.0275 | 16.3934 | 1000 | 2.0282 |
|
| 67 |
-
| 2.0271 | 18.0328 | 1100 | 2.0255 |
|
| 68 |
-
| 2.0205 | 19.6721 | 1200 | 2.0233 |
|
| 69 |
-
| 2.0218 | 21.3115 | 1300 | 2.0219 |
|
| 70 |
-
| 2.011 | 22.9508 | 1400 | 2.0209 |
|
| 71 |
-
|
| 72 |
-
|
| 73 |
### Framework versions
|
| 74 |
|
| 75 |
- PEFT 0.10.0
|
|
|
|
| 6 |
- sft
|
| 7 |
- generated_from_trainer
|
| 8 |
base_model: google/gemma-2b
|
|
|
|
|
|
|
| 9 |
model-index:
|
| 10 |
- name: Not-so-bright-AGI-v1
|
| 11 |
results: []
|
|
|
|
| 16 |
|
| 17 |
# Not-so-bright-AGI-v1
|
| 18 |
|
| 19 |
+
This model is a fine-tuned version of [google/gemma-2b](https://huggingface.co/google/gemma-2b) on an unknown dataset.
|
|
|
|
|
|
|
| 20 |
|
| 21 |
## Model description
|
| 22 |
|
|
|
|
| 36 |
|
| 37 |
The following hyperparameters were used during training:
|
| 38 |
- learning_rate: 1e-05
|
| 39 |
+
- train_batch_size: 4
|
| 40 |
- eval_batch_size: 8
|
| 41 |
- seed: 42
|
| 42 |
- gradient_accumulation_steps: 8
|
| 43 |
+
- total_train_batch_size: 32
|
| 44 |
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
|
| 45 |
- lr_scheduler_type: linear
|
| 46 |
- lr_scheduler_warmup_ratio: 0.05
|
| 47 |
- training_steps: 1480
|
| 48 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 49 |
### Framework versions
|
| 50 |
|
| 51 |
- PEFT 0.10.0
|
adapter_model.safetensors
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 156926880
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:4ac64ba89b8f4a98331f15a564f81eaca49421b8190880d6db4bb35c9a1483d8
|
| 3 |
size 156926880
|