End of training
Browse files- .gitattributes +2 -0
- README.md +8 -1
- adapter_config.json +5 -5
- adapter_model.safetensors +1 -1
- nipsllm_42568.1733596146795479068.pt.trace.json +3 -0
- nipsllm_46661.1733641861005740768.pt.trace.json +3 -0
- runs/Dec07_18-09-36_nipsllm/events.out.tfevents.1733594978.nipsllm.40902.0 +3 -0
- runs/Dec07_18-16-30_nipsllm/events.out.tfevents.1733595391.nipsllm.41938.0 +3 -0
- runs/Dec07_18-23-33_nipsllm/events.out.tfevents.1733595814.nipsllm.42568.0 +3 -0
- runs/Dec08_06-57-21_nipsllm/events.out.tfevents.1733641042.nipsllm.46661.0 +3 -0
- training_args.bin +1 -1
.gitattributes
CHANGED
@@ -38,3 +38,5 @@ nipsllm_13350.1733587985088903357.pt.trace.json filter=lfs diff=lfs merge=lfs -t
|
|
38 |
nipsllm_19414.1733589361760356058.pt.trace.json filter=lfs diff=lfs merge=lfs -text
|
39 |
nipsllm_23043.1733590033744722520.pt.trace.json filter=lfs diff=lfs merge=lfs -text
|
40 |
nipsllm_27141.1733590692262892997.pt.trace.json filter=lfs diff=lfs merge=lfs -text
|
|
|
|
|
|
38 |
nipsllm_19414.1733589361760356058.pt.trace.json filter=lfs diff=lfs merge=lfs -text
|
39 |
nipsllm_23043.1733590033744722520.pt.trace.json filter=lfs diff=lfs merge=lfs -text
|
40 |
nipsllm_27141.1733590692262892997.pt.trace.json filter=lfs diff=lfs merge=lfs -text
|
41 |
+
nipsllm_42568.1733596146795479068.pt.trace.json filter=lfs diff=lfs merge=lfs -text
|
42 |
+
nipsllm_46661.1733641861005740768.pt.trace.json filter=lfs diff=lfs merge=lfs -text
|
README.md
CHANGED
@@ -45,7 +45,7 @@ The following hyperparameters were used during training:
|
|
45 |
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
|
46 |
- lr_scheduler_type: cosine
|
47 |
- lr_scheduler_warmup_ratio: 0.02
|
48 |
-
- num_epochs: 0.
|
49 |
|
50 |
### Training results
|
51 |
|
@@ -87,6 +87,13 @@ The following hyperparameters were used during training:
|
|
87 |
| 1.2623 | 0.4405 | 68 | 1.2332 |
|
88 |
| 1.2845 | 0.4534 | 70 | 1.2331 |
|
89 |
| 1.1966 | 0.4664 | 72 | 1.2331 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
90 |
|
91 |
|
92 |
### Framework versions
|
|
|
45 |
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
|
46 |
- lr_scheduler_type: cosine
|
47 |
- lr_scheduler_warmup_ratio: 0.02
|
48 |
+
- num_epochs: 0.56
|
49 |
|
50 |
### Training results
|
51 |
|
|
|
87 |
| 1.2623 | 0.4405 | 68 | 1.2332 |
|
88 |
| 1.2845 | 0.4534 | 70 | 1.2331 |
|
89 |
| 1.1966 | 0.4664 | 72 | 1.2331 |
|
90 |
+
| 1.2389 | 0.4794 | 74 | 1.2331 |
|
91 |
+
| 1.1957 | 0.4923 | 76 | 1.2331 |
|
92 |
+
| 1.2684 | 0.5053 | 78 | 1.2331 |
|
93 |
+
| 1.3217 | 0.5182 | 80 | 1.2331 |
|
94 |
+
| 1.3126 | 0.5312 | 82 | 1.2331 |
|
95 |
+
| 1.2146 | 0.5441 | 84 | 1.2330 |
|
96 |
+
| 1.216 | 0.5571 | 86 | 1.2331 |
|
97 |
|
98 |
|
99 |
### Framework versions
|
adapter_config.json
CHANGED
@@ -20,13 +20,13 @@
|
|
20 |
"rank_pattern": {},
|
21 |
"revision": null,
|
22 |
"target_modules": [
|
|
|
|
|
|
|
23 |
"gate_proj",
|
24 |
-
"o_proj",
|
25 |
"k_proj",
|
26 |
-
"
|
27 |
-
"
|
28 |
-
"up_proj",
|
29 |
-
"q_proj"
|
30 |
],
|
31 |
"task_type": "CAUSAL_LM",
|
32 |
"use_dora": false,
|
|
|
20 |
"rank_pattern": {},
|
21 |
"revision": null,
|
22 |
"target_modules": [
|
23 |
+
"q_proj",
|
24 |
+
"down_proj",
|
25 |
+
"v_proj",
|
26 |
"gate_proj",
|
|
|
27 |
"k_proj",
|
28 |
+
"o_proj",
|
29 |
+
"up_proj"
|
|
|
|
|
30 |
],
|
31 |
"task_type": "CAUSAL_LM",
|
32 |
"use_dora": false,
|
adapter_model.safetensors
CHANGED
@@ -1,3 +1,3 @@
|
|
1 |
version https://git-lfs.github.com/spec/v1
|
2 |
-
oid sha256:
|
3 |
size 83945296
|
|
|
1 |
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:c6acf9c2eda2a5947a687b68b6a88c402f84e30473d721be6d2ac0dd94de6b6c
|
3 |
size 83945296
|
nipsllm_42568.1733596146795479068.pt.trace.json
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:78f13e0a1f2750a9f117ba2f1a6d0682b9b218212a491b674e1d2702c5c036e8
|
3 |
+
size 1881282838
|
nipsllm_46661.1733641861005740768.pt.trace.json
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:f03ef86a80810b17ae89c1278332bdead6a864d6ba157e7aa33581ede2835882
|
3 |
+
size 4444007570
|
runs/Dec07_18-09-36_nipsllm/events.out.tfevents.1733594978.nipsllm.40902.0
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:6213c8aa2f31255ff15f720afb1503782b73f5caa574548d76c0d0fa77d4bad5
|
3 |
+
size 6216
|
runs/Dec07_18-16-30_nipsllm/events.out.tfevents.1733595391.nipsllm.41938.0
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:27ab31ad97254a8e3fd976c2d03e4bb70df0a6259447bc16a6d8b9327a1e0d22
|
3 |
+
size 6423
|
runs/Dec07_18-23-33_nipsllm/events.out.tfevents.1733595814.nipsllm.42568.0
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:f686db6d67aa1447acb9989e42851f68c309cd35387563ca82784730d07b0d9c
|
3 |
+
size 6423
|
runs/Dec08_06-57-21_nipsllm/events.out.tfevents.1733641042.nipsllm.46661.0
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:2f97bfbbc21325ea8e47c221ef2a565270b60d670cbf3d008adfdc93efc4016a
|
3 |
+
size 8601
|
training_args.bin
CHANGED
@@ -1,3 +1,3 @@
|
|
1 |
version https://git-lfs.github.com/spec/v1
|
2 |
-
oid sha256:
|
3 |
size 5304
|
|
|
1 |
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:fa4424a3d85efd28afc863043bf02d1dacb8ce68e6b4f55c89131718a1561ab8
|
3 |
size 5304
|