DavidAU
/

Qwen2.5-3X7B-CoderInstruct-OlympicCoder-MS-Next-Coder-25B-v1-128k-ctx

Model card Files Files and versions Community

DavidAU commited on Jul 14

Commit

9683cf3

·

verified ·

1 Parent(s): 0f6525f

Update README.md

Files changed (1) hide show

README.md +2 -2

README.md CHANGED Viewed

@@ -26,7 +26,7 @@ tags:
 library_name: transformers
 ---
-<h2>Qwen2.5-3X7B-CoderInstruct-OlympicCoder-MS-Next-Coder-25B-v1</h2>
 This repo contains the full precision source code, in "safe tensors" format to generate GGUFs, GPTQ, EXL2, AWQ, HQQ and other formats.
 The source code can also be used directly.
@@ -46,7 +46,7 @@ Suggest using 3 experts for maximum power.
 SETTINGS:
 - Temp .5 to .7 (or lower)
-- Max Context is 32k (final versions will be 32k, 128k and maybe higher).
 - topk: 20, topp: .8, minp: .05
 - rep pen: 1.05-1.1 (can be lower)
 - Jinja Template (embedded) or CHATML template.

 library_name: transformers
 ---
+<h2>Qwen2.5-3X7B-CoderInstruct-OlympicCoder-MS-Next-Coder-25B-v1-128k-ctx</h2>
 This repo contains the full precision source code, in "safe tensors" format to generate GGUFs, GPTQ, EXL2, AWQ, HQQ and other formats.
 The source code can also be used directly.
 SETTINGS:
 - Temp .5 to .7 (or lower)
+- Max Context is 128k via YARN as per Qwen specs.
 - topk: 20, topp: .8, minp: .05
 - rep pen: 1.05-1.1 (can be lower)
 - Jinja Template (embedded) or CHATML template.