Model save
Browse files- README.md +6 -5
- adapter_config.json +3 -3
- adapter_model.safetensors +1 -1
- runs/Jun04_19-40-55_702dae2383a0/events.out.tfevents.1717530079.702dae2383a0.121172.0 +2 -2
- step_0/README.md +0 -77
- step_0/adapter_config.json +3 -3
- step_0/training_args.bin +1 -1
- step_3600/adapter_config.json +3 -3
- step_3600/adapter_model.safetensors +1 -1
- step_3600/training_args.bin +1 -1
- training_args.bin +1 -1
README.md
CHANGED
@@ -16,7 +16,7 @@ model-index:
|
|
16 |
<!-- This model card has been generated automatically according to the information the Trainer had access to. You
|
17 |
should probably proofread and complete it, then remove this comment. -->
|
18 |
|
19 |
-
[<img src="https://raw.githubusercontent.com/wandb/assets/main/wandb-github-badge-28.svg" alt="Visualize in Weights & Biases" width="200" height="32"/>](https://wandb.ai/quirky_lats_at_mats/harmbench_adv_training/runs/
|
20 |
# llama2-7b-sft-lora
|
21 |
|
22 |
This model is a fine-tuned version of [meta-llama/Llama-2-7b-chat-hf](https://huggingface.co/meta-llama/Llama-2-7b-chat-hf) on the generator dataset.
|
@@ -38,14 +38,15 @@ More information needed
|
|
38 |
### Training hyperparameters
|
39 |
|
40 |
The following hyperparameters were used during training:
|
41 |
-
- learning_rate:
|
42 |
- train_batch_size: 4
|
43 |
- eval_batch_size: 8
|
44 |
- seed: 4
|
45 |
- distributed_type: multi-GPU
|
46 |
-
- num_devices:
|
47 |
-
-
|
48 |
-
-
|
|
|
49 |
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
|
50 |
- lr_scheduler_type: cosine
|
51 |
- training_steps: 2000
|
|
|
16 |
<!-- This model card has been generated automatically according to the information the Trainer had access to. You
|
17 |
should probably proofread and complete it, then remove this comment. -->
|
18 |
|
19 |
+
[<img src="https://raw.githubusercontent.com/wandb/assets/main/wandb-github-badge-28.svg" alt="Visualize in Weights & Biases" width="200" height="32"/>](https://wandb.ai/quirky_lats_at_mats/harmbench_adv_training/runs/mgxwn0ra)
|
20 |
# llama2-7b-sft-lora
|
21 |
|
22 |
This model is a fine-tuned version of [meta-llama/Llama-2-7b-chat-hf](https://huggingface.co/meta-llama/Llama-2-7b-chat-hf) on the generator dataset.
|
|
|
38 |
### Training hyperparameters
|
39 |
|
40 |
The following hyperparameters were used during training:
|
41 |
+
- learning_rate: 2e-05
|
42 |
- train_batch_size: 4
|
43 |
- eval_batch_size: 8
|
44 |
- seed: 4
|
45 |
- distributed_type: multi-GPU
|
46 |
+
- num_devices: 8
|
47 |
+
- gradient_accumulation_steps: 4
|
48 |
+
- total_train_batch_size: 128
|
49 |
+
- total_eval_batch_size: 64
|
50 |
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
|
51 |
- lr_scheduler_type: cosine
|
52 |
- training_steps: 2000
|
adapter_config.json
CHANGED
@@ -16,12 +16,12 @@
|
|
16 |
"rank_pattern": {},
|
17 |
"revision": null,
|
18 |
"target_modules": [
|
|
|
19 |
"down_proj",
|
20 |
-
"
|
21 |
"q_proj",
|
22 |
"up_proj",
|
23 |
-
"
|
24 |
-
"v_proj"
|
25 |
],
|
26 |
"task_type": "CAUSAL_LM"
|
27 |
}
|
|
|
16 |
"rank_pattern": {},
|
17 |
"revision": null,
|
18 |
"target_modules": [
|
19 |
+
"k_proj",
|
20 |
"down_proj",
|
21 |
+
"v_proj",
|
22 |
"q_proj",
|
23 |
"up_proj",
|
24 |
+
"o_proj"
|
|
|
25 |
],
|
26 |
"task_type": "CAUSAL_LM"
|
27 |
}
|
adapter_model.safetensors
CHANGED
@@ -1,3 +1,3 @@
|
|
1 |
version https://git-lfs.github.com/spec/v1
|
2 |
-
oid sha256:
|
3 |
size 258001832
|
|
|
1 |
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:6101862afadc009beafc9118e8389868e023587268631adb4d7c1da2b232bc16
|
3 |
size 258001832
|
runs/Jun04_19-40-55_702dae2383a0/events.out.tfevents.1717530079.702dae2383a0.121172.0
CHANGED
@@ -1,3 +1,3 @@
|
|
1 |
version https://git-lfs.github.com/spec/v1
|
2 |
-
oid sha256:
|
3 |
-
size
|
|
|
1 |
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:a4b77558a253998ce7bc7f631204ca506ada016bf05766a9c4d7e394b8167aa3
|
3 |
+
size 145988
|
step_0/README.md
CHANGED
@@ -266,81 +266,4 @@ Carbon emissions can be estimated using the [Machine Learning Impact calculator]
|
|
266 |
### Framework versions
|
267 |
|
268 |
|
269 |
-
- PEFT 0.6.1
|
270 |
-
## Training procedure
|
271 |
-
|
272 |
-
|
273 |
-
### Framework versions
|
274 |
-
|
275 |
-
|
276 |
-
- PEFT 0.6.1
|
277 |
-
## Training procedure
|
278 |
-
|
279 |
-
|
280 |
-
### Framework versions
|
281 |
-
|
282 |
-
|
283 |
-
- PEFT 0.6.1
|
284 |
-
## Training procedure
|
285 |
-
|
286 |
-
|
287 |
-
### Framework versions
|
288 |
-
|
289 |
-
|
290 |
-
- PEFT 0.6.1
|
291 |
-
## Training procedure
|
292 |
-
|
293 |
-
|
294 |
-
### Framework versions
|
295 |
-
|
296 |
-
|
297 |
-
- PEFT 0.6.1
|
298 |
-
## Training procedure
|
299 |
-
|
300 |
-
|
301 |
-
### Framework versions
|
302 |
-
|
303 |
-
|
304 |
-
- PEFT 0.6.1
|
305 |
-
## Training procedure
|
306 |
-
|
307 |
-
|
308 |
-
### Framework versions
|
309 |
-
|
310 |
-
|
311 |
-
- PEFT 0.6.1
|
312 |
-
## Training procedure
|
313 |
-
|
314 |
-
|
315 |
-
### Framework versions
|
316 |
-
|
317 |
-
|
318 |
-
- PEFT 0.6.1
|
319 |
-
## Training procedure
|
320 |
-
|
321 |
-
|
322 |
-
### Framework versions
|
323 |
-
|
324 |
-
|
325 |
-
- PEFT 0.6.1
|
326 |
-
## Training procedure
|
327 |
-
|
328 |
-
|
329 |
-
### Framework versions
|
330 |
-
|
331 |
-
|
332 |
-
- PEFT 0.6.1
|
333 |
-
## Training procedure
|
334 |
-
|
335 |
-
|
336 |
-
### Framework versions
|
337 |
-
|
338 |
-
|
339 |
-
- PEFT 0.6.1
|
340 |
-
## Training procedure
|
341 |
-
|
342 |
-
|
343 |
-
### Framework versions
|
344 |
-
|
345 |
-
|
346 |
- PEFT 0.6.1
|
|
|
266 |
### Framework versions
|
267 |
|
268 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
269 |
- PEFT 0.6.1
|
step_0/adapter_config.json
CHANGED
@@ -16,12 +16,12 @@
|
|
16 |
"rank_pattern": {},
|
17 |
"revision": null,
|
18 |
"target_modules": [
|
|
|
19 |
"down_proj",
|
20 |
-
"
|
21 |
"q_proj",
|
22 |
"up_proj",
|
23 |
-
"
|
24 |
-
"v_proj"
|
25 |
],
|
26 |
"task_type": "CAUSAL_LM"
|
27 |
}
|
|
|
16 |
"rank_pattern": {},
|
17 |
"revision": null,
|
18 |
"target_modules": [
|
19 |
+
"k_proj",
|
20 |
"down_proj",
|
21 |
+
"v_proj",
|
22 |
"q_proj",
|
23 |
"up_proj",
|
24 |
+
"o_proj"
|
|
|
25 |
],
|
26 |
"task_type": "CAUSAL_LM"
|
27 |
}
|
step_0/training_args.bin
CHANGED
@@ -1,3 +1,3 @@
|
|
1 |
version https://git-lfs.github.com/spec/v1
|
2 |
-
oid sha256:
|
3 |
size 6200
|
|
|
1 |
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:fddcb958cb044b08eb2784936cc2d61c7603663890e064c856baab404ea4a540
|
3 |
size 6200
|
step_3600/adapter_config.json
CHANGED
@@ -16,12 +16,12 @@
|
|
16 |
"rank_pattern": {},
|
17 |
"revision": null,
|
18 |
"target_modules": [
|
|
|
|
|
19 |
"v_proj",
|
20 |
-
"o_proj",
|
21 |
"q_proj",
|
22 |
"up_proj",
|
23 |
-
"
|
24 |
-
"k_proj"
|
25 |
],
|
26 |
"task_type": "CAUSAL_LM"
|
27 |
}
|
|
|
16 |
"rank_pattern": {},
|
17 |
"revision": null,
|
18 |
"target_modules": [
|
19 |
+
"k_proj",
|
20 |
+
"down_proj",
|
21 |
"v_proj",
|
|
|
22 |
"q_proj",
|
23 |
"up_proj",
|
24 |
+
"o_proj"
|
|
|
25 |
],
|
26 |
"task_type": "CAUSAL_LM"
|
27 |
}
|
step_3600/adapter_model.safetensors
CHANGED
@@ -1,3 +1,3 @@
|
|
1 |
version https://git-lfs.github.com/spec/v1
|
2 |
-
oid sha256:
|
3 |
size 258001832
|
|
|
1 |
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:6101862afadc009beafc9118e8389868e023587268631adb4d7c1da2b232bc16
|
3 |
size 258001832
|
step_3600/training_args.bin
CHANGED
@@ -1,3 +1,3 @@
|
|
1 |
version https://git-lfs.github.com/spec/v1
|
2 |
-
oid sha256:
|
3 |
size 6200
|
|
|
1 |
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:fddcb958cb044b08eb2784936cc2d61c7603663890e064c856baab404ea4a540
|
3 |
size 6200
|
training_args.bin
CHANGED
@@ -1,3 +1,3 @@
|
|
1 |
version https://git-lfs.github.com/spec/v1
|
2 |
-
oid sha256:
|
3 |
size 6200
|
|
|
1 |
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:fddcb958cb044b08eb2784936cc2d61c7603663890e064c856baab404ea4a540
|
3 |
size 6200
|