Add razvilka evaluation result
#4
by ai-forever - opened
YAML Metadata Error:Invalid content in Eval Result file .eval_results/razvilka.yaml
Check out the documentation for more information.
Show details
Task ID "accuracy" does not match any task in dataset "artemsnegirev/razvilka". Available: none
.eval_results/razvilka.yaml
ADDED
|
@@ -0,0 +1,9 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
- dataset:
|
| 2 |
+
id: artemsnegirev/razvilka
|
| 3 |
+
task_id: accuracy
|
| 4 |
+
revision: a31c9404615592085e81f971c65896ab91afd179
|
| 5 |
+
value: 0.893
|
| 6 |
+
source:
|
| 7 |
+
url: https://github.com/ai-forever/FRIDA-Decisions/tree/v0.4.0/benchmarks/razvilka
|
| 8 |
+
name: razvilka_eval.py
|
| 9 |
+
notes: "PyTorch bf16, state budget 512 tokens; int8 ONNX on CPU 0.891"
|