philtheriver commited on
Commit
9b3d64e
Β·
verified Β·
1 Parent(s): f1a2fdf

Update README.md

Browse files
Files changed (1) hide show
  1. README.md +1 -2
README.md CHANGED
@@ -32,9 +32,8 @@ ROCmFPX 3β†’8-bit quants of [`Jackrong/Qwopus3.6-27B-Coder-MTP-GGUF`](https://hu
32
  |---|---|
33
  | **Quality** | β‰ˆ stock `Q6_K` β€” PPL **+0.17%** (within error) |
34
  | **Prompt processing** | **+32%** vs Q6_K (short ctx) β†’ +22% at 64k |
35
- | **Decode** | ~18 tok/s with MTP (~9 raw) |
36
  | **Vision** | Qwen3-VL β€” bundled `mmproj/` |
37
- | **Context** | up to 128k |
38
 
39
  > ⚠️ **Requires the [ROCmFPX fork](https://github.com/charlie12345/ROCmFPX) @ [`553504b`](https://github.com/charlie12345/ROCmFPX/commit/553504bd5d16fe1374e9b1508a1f395c9a35d746)** β€” custom AMD quant types (enum IDs 110–115), not upstream-stable. **Won't load** in stock llama.cpp / LM Studio / Ollama. HF's precision badge is wrong β€” **pick the file by name**.
40
 
 
32
  |---|---|
33
  | **Quality** | β‰ˆ stock `Q6_K` β€” PPL **+0.17%** (within error) |
34
  | **Prompt processing** | **+32%** vs Q6_K (short ctx) β†’ +22% at 64k |
35
+ | **Decode** | \~18 tok/s with MTP (~9 raw) |
36
  | **Vision** | Qwen3-VL β€” bundled `mmproj/` |
 
37
 
38
  > ⚠️ **Requires the [ROCmFPX fork](https://github.com/charlie12345/ROCmFPX) @ [`553504b`](https://github.com/charlie12345/ROCmFPX/commit/553504bd5d16fe1374e9b1508a1f395c9a35d746)** β€” custom AMD quant types (enum IDs 110–115), not upstream-stable. **Won't load** in stock llama.cpp / LM Studio / Ollama. HF's precision badge is wrong β€” **pick the file by name**.
39