Update model card with benchmark data source
Browse files
README.md
CHANGED
|
@@ -12,8 +12,10 @@ model-index:
|
|
| 12 |
- task:
|
| 13 |
type: math-evaluation
|
| 14 |
dataset:
|
| 15 |
-
type:
|
| 16 |
name: OpenMathInstruct
|
|
|
|
|
|
|
| 17 |
metrics:
|
| 18 |
- name: exact_match,none
|
| 19 |
type: exact_match
|
|
@@ -40,6 +42,8 @@ model-index:
|
|
| 40 |
dataset:
|
| 41 |
type: meta/arc-dataset
|
| 42 |
name: Meta-ARC Dataset
|
|
|
|
|
|
|
| 43 |
metrics:
|
| 44 |
- name: exact_match,strict-match
|
| 45 |
type: exact_match
|
|
@@ -68,4 +72,4 @@ model-index:
|
|
| 68 |
verified: false
|
| 69 |
---
|
| 70 |
# Control-LLM-Llama3.1-8B-Math16
|
| 71 |
-
This is a fine-tuned model of Llama-3.1-8B-Instruct for mathematical tasks on OpenMath2 dataset.
|
|
|
|
| 12 |
- task:
|
| 13 |
type: math-evaluation
|
| 14 |
dataset:
|
| 15 |
+
type: parquet
|
| 16 |
name: OpenMathInstruct
|
| 17 |
+
dataset_kwargs:
|
| 18 |
+
data_files: "/home/jobuser/controlllm/inference/llm_eval_harness/additional_tasks/math/joined_math.parquet"
|
| 19 |
metrics:
|
| 20 |
- name: exact_match,none
|
| 21 |
type: exact_match
|
|
|
|
| 42 |
dataset:
|
| 43 |
type: meta/arc-dataset
|
| 44 |
name: Meta-ARC Dataset
|
| 45 |
+
dataset_path: "meta-llama/llama-3.1-8_b-instruct-evals"
|
| 46 |
+
dataset_name: "Llama-3.1-8B-Instruct-evals__arc_challenge__details"
|
| 47 |
metrics:
|
| 48 |
- name: exact_match,strict-match
|
| 49 |
type: exact_match
|
|
|
|
| 72 |
verified: false
|
| 73 |
---
|
| 74 |
# Control-LLM-Llama3.1-8B-Math16
|
| 75 |
+
This is a fine-tuned model of Llama-3.1-8B-Instruct for mathematical tasks on OpenMath2 dataset.
|