Update README.md
Browse files
README.md
CHANGED
|
@@ -6,6 +6,10 @@ metrics:
|
|
| 6 |
- accuracy
|
| 7 |
language:
|
| 8 |
- en
|
|
|
|
|
|
|
|
|
|
|
|
|
| 9 |
---
|
| 10 |
|
| 11 |
|
|
@@ -13,8 +17,9 @@ language:
|
|
| 13 |
|
| 14 |
## Model description
|
| 15 |
|
| 16 |
-
|
| 17 |
-
|
|
|
|
| 18 |
- Loss: 1.9926
|
| 19 |
- Accuracy: 0.5885
|
| 20 |
- Num Input Tokens Seen: 3468165120
|
|
@@ -59,38 +64,3 @@ The following hyperparameters were used during training:
|
|
| 59 |
- lr_scheduler_type: linear
|
| 60 |
- lr_scheduler_warmup_ratio: 0.05
|
| 61 |
- num_epochs: 1.0
|
| 62 |
-
|
| 63 |
-
### Training results
|
| 64 |
-
|
| 65 |
-
| Training Loss | Epoch | Step | Validation Loss | Accuracy | Input Tokens Seen |
|
| 66 |
-
|:-------------:|:------:|:----:|:---------------:|:--------:|:-----------------:|
|
| 67 |
-
| 2.2374 | 0.0454 | 400 | 2.1871 | 0.5588 | 157286400 |
|
| 68 |
-
| 2.143 | 0.0907 | 800 | 2.1336 | 0.5665 | 314572800 |
|
| 69 |
-
| 2.1272 | 0.1361 | 1200 | 2.1092 | 0.5698 | 471859200 |
|
| 70 |
-
| 2.1243 | 0.1814 | 1600 | 2.0929 | 0.5725 | 629145600 |
|
| 71 |
-
| 2.1021 | 0.2268 | 2000 | 2.0794 | 0.5747 | 786432000 |
|
| 72 |
-
| 2.0794 | 0.2721 | 2400 | 2.0687 | 0.5762 | 943718400 |
|
| 73 |
-
| 2.0843 | 0.3175 | 2800 | 2.0592 | 0.5776 | 1101004800 |
|
| 74 |
-
| 2.0571 | 0.3628 | 3200 | 2.0507 | 0.5793 | 1258291200 |
|
| 75 |
-
| 2.0841 | 0.4082 | 3600 | 2.0435 | 0.5802 | 1415577600 |
|
| 76 |
-
| 2.0484 | 0.4535 | 4000 | 2.0363 | 0.5813 | 1572864000 |
|
| 77 |
-
| 2.0199 | 0.4989 | 4400 | 2.0315 | 0.5820 | 1730150400 |
|
| 78 |
-
| 2.0361 | 0.5442 | 4800 | 2.0261 | 0.5829 | 1887436800 |
|
| 79 |
-
| 2.057 | 0.5896 | 5200 | 2.0207 | 0.5838 | 2044723200 |
|
| 80 |
-
| 2.0234 | 0.6349 | 5600 | 2.0163 | 0.5845 | 2202009600 |
|
| 81 |
-
| 2.073 | 0.6803 | 6000 | 2.0120 | 0.5850 | 2359296000 |
|
| 82 |
-
| 2.058 | 0.7256 | 6400 | 2.0074 | 0.5862 | 2516582400 |
|
| 83 |
-
| 2.0253 | 0.7710 | 6800 | 2.0041 | 0.5866 | 2673868800 |
|
| 84 |
-
| 1.995 | 0.8163 | 7200 | 2.0010 | 0.5872 | 2831155200 |
|
| 85 |
-
| 1.9735 | 0.8617 | 7600 | 1.9987 | 0.5875 | 2988441600 |
|
| 86 |
-
| 1.9799 | 0.9070 | 8000 | 1.9960 | 0.5880 | 3145728000 |
|
| 87 |
-
| 2.0056 | 0.9524 | 8400 | 1.9942 | 0.5882 | 3303014400 |
|
| 88 |
-
| 1.9961 | 0.9977 | 8800 | 1.9926 | 0.5884 | 3460300800 |
|
| 89 |
-
|
| 90 |
-
|
| 91 |
-
### Framework versions
|
| 92 |
-
|
| 93 |
-
- Transformers 4.40.2
|
| 94 |
-
- Pytorch 2.2.0+cu121
|
| 95 |
-
- Datasets 2.19.1
|
| 96 |
-
- Tokenizers 0.19.1
|
|
|
|
| 6 |
- accuracy
|
| 7 |
language:
|
| 8 |
- en
|
| 9 |
+
datasets:
|
| 10 |
+
- BEE-spoke-data/UltraTextbooks-2.1-fw_mix
|
| 11 |
+
- BEE-spoke-data/napierone-epub-raw
|
| 12 |
+
- BEE-spoke-data/knowledge-inoc-concat-v1
|
| 13 |
---
|
| 14 |
|
| 15 |
|
|
|
|
| 17 |
|
| 18 |
## Model description
|
| 19 |
|
| 20 |
+
Continued-training of [BEE-spoke-data/mega-ar-350m-L3t-v0.08-ultraTBfw](https://hf.co/BEE-spoke-data/mega-ar-350m-L3t-v0.08-ultraTBfw) on a few more datasets.
|
| 21 |
+
|
| 22 |
+
It achieves the following results on the evaluation set (`BEE-spoke-data/UltraTextbooks-2.1-fw_mix`):
|
| 23 |
- Loss: 1.9926
|
| 24 |
- Accuracy: 0.5885
|
| 25 |
- Num Input Tokens Seen: 3468165120
|
|
|
|
| 64 |
- lr_scheduler_type: linear
|
| 65 |
- lr_scheduler_warmup_ratio: 0.05
|
| 66 |
- num_epochs: 1.0
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|