sethuiyer commited on
Commit
23399ca
Β·
verified Β·
1 Parent(s): 8cb8f49

Update README.md

Browse files
Files changed (1) hide show
  1. README.md +6 -5
README.md CHANGED
@@ -99,12 +99,13 @@ model-index:
99
  ---
100
  # Llama 3.1 8B Experimental 1206
101
 
102
- Overall Strengths
103
- Logical and Boolean Reasoning – Excels in tasks requiring clear, rule-based logic and manipulation of true/false statements.
104
- Focused Domain Knowledge – Strong at certain specialized tasks (sports rules, ruin names, hyperbaton) that blend world knowledge with language comprehension.
105
- Good Instruction Compliance – High prompt-level and instance-level accuracy (both strict and loose) indicate it follows user instructions effectively.
106
- Reasonable Multi-step Reasoning – While not top-tier across all logic tasks, it still shows solid performance in areas like disambiguation and causal reasoning (over 60%).
107
 
 
 
 
 
 
108
 
109
  # [Open LLM Leaderboard Evaluation Results](https://huggingface.co/spaces/open-llm-leaderboard/open_llm_leaderboard)
110
  Detailed results can be found [here](https://huggingface.co/datasets/open-llm-leaderboard/sethuiyer__Llama-3.1-8B-Experimental-1206-Instruct-details)
 
99
  ---
100
  # Llama 3.1 8B Experimental 1206
101
 
102
+ ### Overall Strengths
 
 
 
 
103
 
104
+ 1. **Logical and Boolean Reasoning** – Excels in tasks requiring clear, rule-based logic and manipulation of true/false statements.
105
+ 2. **Focused Domain Knowledge** – Strong at certain specialized tasks (sports rules, ruin names, hyperbaton) that blend world knowledge with language comprehension.
106
+ 3. **Good Instruction Compliance** – High prompt-level and instance-level accuracy (both strict and loose) indicate that it follows user instructions effectively, even in more complex or nuanced prompts.
107
+ 4. **Reasonable Multi-step Reasoning** – While not the best in every logic category, it still shows solid performance (60%+) on tasks like disambiguation and causal reasoning.
108
+ 5. **Extended Context Window (138k)** – The large 138k token context allows the model to handle lengthy inputs and maintain coherence across extensive passages or multi-turn conversations. This is especially valuable for tasks like long-document question answering, summarization, or complex scenario analysis where context retention is crucial.
109
 
110
  # [Open LLM Leaderboard Evaluation Results](https://huggingface.co/spaces/open-llm-leaderboard/open_llm_leaderboard)
111
  Detailed results can be found [here](https://huggingface.co/datasets/open-llm-leaderboard/sethuiyer__Llama-3.1-8B-Experimental-1206-Instruct-details)