Multi-turn Conversation Evaluation on IMDb
5.6MT-Bench ScoreCOALA
Evaluation Results
| Method | Links | |
|---|---|---|
| COALABase Model=DOLPHIN-7B2026.05 | 5.6 | |
| ORPOBase Model=DOLPHIN-7B2026.05 | 5.5 | |
| DPOBase Model=DOLPHIN-7B2026.05 | 5.1 | |
| SFTBase Model=LLAMA-8B2026.05 | 3.5 | |
| COALABase Model=LLAMA-8B2026.05 | 3.5 | |
| SFTBase Model=DOLPHIN-7B2026.05 | 3.2 | |
| DPOBase Model=LLAMA-8B2026.05 | 3.1 | |
| ORPOBase Model=LLAMA-8B2026.05 | 2.9 | |
| COALABase Model=MISTRAL-7B2026.05 | 2.9 | |
| DPOBase Model=MISTRAL-7B2026.05 | 2.8 | |
| SFTBase Model=MISTRAL-7B2026.05 | 2.3 | |
| ORPOBase Model=MISTRAL-7B2026.05 | 1.7 | |
| DPOBase Model=DISTILGPT2026.05 | 1.6 | |
| SFTBase Model=GPT-22026.05 | 1.4 | |
| SFTBase Model=DISTILGPT2026.05 | 1.2 | |
| ORPOBase Model=DISTILGPT2026.05 | 1.2 | |
| COALABase Model=DISTILGPT2026.05 | 1.2 | |
| ORPOBase Model=GPT-22026.05 | 1.1 | |
| COALABase Model=GPT-22026.05 | 1.1 | |
| DPOBase Model=GPT-22026.05 | 1 |