Create lm_harness_results.md
Browse files- lm_harness_results.md +21 -0
lm_harness_results.md
ADDED
|
@@ -0,0 +1,21 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
## Quick eval
|
| 2 |
+
|
| 3 |
+
Quick eval for: `BEE-spoke-data/tFINE-900m-e16-d32-instruct_2e`
|
| 4 |
+
|
| 5 |
+
|
| 6 |
+
hf (pretrained=BEE-spoke-data/tFINE-900m-e16-d32-instruct_2e,trust_remote_code=True,dtype=bfloat16,trust_remote_code=True), gen_kwargs: (None), limit: None, num_fewshot: None, batch_size: 8
|
| 7 |
+
|
| 8 |
+
| Tasks |Version| Filter |n-shot| Metric | |Value | |Stderr|
|
| 9 |
+
|-------------|------:|----------------|-----:|-----------|---|-----:|---|------|
|
| 10 |
+
|boolq | 2|none | 0|acc |↑ |0.6254|± |0.0085|
|
| 11 |
+
|openbookqa | 1|none | 0|acc |↑ |0.1520|± |0.0161|
|
| 12 |
+
| | |none | 0|acc_norm |↑ |0.3100|± |0.0207|
|
| 13 |
+
|piqa | 1|none | 0|acc |↑ |0.6072|± |0.0114|
|
| 14 |
+
| | |none | 0|acc_norm |↑ |0.5996|± |0.0114|
|
| 15 |
+
|social_iqa | 0|none | 0|acc |↑ |0.4212|± |0.0112|
|
| 16 |
+
|tinyArc | 0|none | 25|acc_norm |↑ |0.2998|± | N/A|
|
| 17 |
+
|tinyGSM8k | 0|flexible-extract| 5|exact_match|↑ |0.0605|± | N/A|
|
| 18 |
+
| | |strict-match | 5|exact_match|↑ |0.0432|± | N/A|
|
| 19 |
+
|tinyHellaswag| 0|none | 10|acc_norm |↑ |0.2969|± | N/A|
|
| 20 |
+
|tinyMMLU | 0|none | 0|acc_norm |↑ |0.3120|± | N/A|
|
| 21 |
+
|winogrande | 1|none | 0|acc |↑ |0.4964|± |0.0141|
|