Model card: long-context acceptance (AA-LCR, BEAM 128K, BEAM 1M)
Browse files
README.md
CHANGED
|
@@ -40,3 +40,14 @@ DSpark draft model for Kimi K3.1.
|
|
| 40 |
| SWE-bench Pro | 642 | 2,048 | 3.379 | 0.77 0.56 0.39 0.27 0.19 0.12 0.08 |
|
| 41 |
| AIME 2026 | 30 | 4,096 | 2.955 | 0.71 0.47 0.31 0.20 0.13 0.08 0.05 |
|
| 42 |
| **Pooled** | | | **4.035** | 0.82 0.64 0.50 0.38 0.30 0.23 0.18 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 40 |
| SWE-bench Pro | 642 | 2,048 | 3.379 | 0.77 0.56 0.39 0.27 0.19 0.12 0.08 |
|
| 41 |
| AIME 2026 | 30 | 4,096 | 2.955 | 0.71 0.47 0.31 0.20 0.13 0.08 0.05 |
|
| 42 |
| **Pooled** | | | **4.035** | 0.82 0.64 0.50 0.38 0.30 0.23 0.18 |
|
| 43 |
+
|
| 44 |
+
## Long context
|
| 45 |
+
|
| 46 |
+
Same settings with prefix caching on. AA-LCR sends its assembled prompts as completions; each BEAM
|
| 47 |
+
question follows its full conversation as a chat completion.
|
| 48 |
+
|
| 49 |
+
| Benchmark | Prompts | Prompt tokens | Max tokens | Acceptance length | Accept rate, positions 1-7 |
|
| 50 |
+
| --- | ---: | ---: | ---: | ---: | --- |
|
| 51 |
+
| AA-LCR | 100 | 95K mean | 1,024 | **3.893** | 0.80 0.63 0.49 0.37 0.27 0.20 0.14 |
|
| 52 |
+
| BEAM 128K | 400 | 97K-125K | 2,048 | 3.809 | 0.82 0.63 0.47 0.35 0.25 0.17 0.12 |
|
| 53 |
+
| BEAM 1M | 700 | 826K-1.04M | 2,048 | 3.255 | 0.78 0.55 0.38 0.24 0.15 0.09 0.06 |
|