Update README.md

README.md

@@ -48,23 +48,23 @@ This training run is monolingual and uses c4en and english wikipedia datasets.
 
 ## Test results
 
-These are the results from [EleutherAI/lm-evaluation-harness](https://github.com/EleutherAI/lm-evaluation-harness) at 39B (tokens trained) checkpoint.
+These are the results from [EleutherAI/lm-evaluation-harness](https://github.com/EleutherAI/lm-evaluation-harness) at 50B (tokens trained) checkpoint.
 
 |     Task     |Version| Metric | Value |   |Stderr|
 |--------------|------:|--------|------:|---|-----:|
-|anli_r1       |      0|acc     | 0.3390|±  |0.0150|
-|anli_r2       |      0|acc     | 0.3350|±  |0.0149|
-|anli_r3       |      0|acc     | 0.3400|±  |0.0137|
-|hellaswag     |      0|acc     | 0.4332|±  |0.0049|
-|              |       |acc_norm| 0.5628|±  |0.0050|
-|lambada_openai|      0|ppl     |13.2084|±  |0.4599|
-|              |       |acc     | 0.4890|±  |0.0070|
-|mathqa        |      0|acc     | 0.2235|±  |0.0076|
-|              |       |acc_norm| 0.2275|±  |0.0077|
-|piqa          |      0|acc     | 0.7361|±  |0.0103|
-|              |       |acc_norm| 0.7399|±  |0.0102|
-|winogrande    |      0|acc     | 0.5596|±  |0.0140|
-|wsc           |      0|acc     | 0.3942|±  |0.0482|
+|anli_r1       |      0|acc     | 0.3480|±  |0.0151|
+|anli_r2       |      0|acc     | 0.3340|±  |0.0149|
+|anli_r3       |      0|acc     | 0.3375|±  |0.0137|
+|hellaswag     |      0|acc     | 0.4476|±  |0.0050|
+|              |       |acc_norm| 0.5904|±  |0.0049|
+|lambada_openai|      0|ppl     |11.0912|±  |0.3672|
+|              |       |acc     | 0.5257|±  |0.0070|
+|mathqa        |      0|acc     | 0.2315|±  |0.0077|
+|              |       |acc_norm| 0.2318|±  |0.0077|
+|piqa          |      0|acc     | 0.7546|±  |0.0100|
+|              |       |acc_norm| 0.7481|±  |0.0101|
+|winogrande    |      0|acc     | 0.5754|±  |0.0139|
+|wsc           |      0|acc     | 0.5000|±  |0.0493|
 
 
 ## Installation