github-starred/ollama
2
0
Fork 0
mirror of https://github.com/ollama/ollama.git synced 2026-05-06 16:11:34 -05:00
Code Issues 14.5k Packages Projects Releases 120 Wiki Activity

[GH-ISSUE #12354] Updated qwen3-coder model shows incorrect quantization #33967

New Issue
Closed
opened 2026-04-22 17:09:39 -05:00 by GiteaMirror · 2 comments
GiteaMirror commented 2026-04-22 17:09:39 -05:00
Owner
Copy Link

Originally created by @rick-github on GitHub (Sep 20, 2025).
Original GitHub issue: https://github.com/ollama/ollama/issues/12354

What is the issue?

The update for tool support has changed the quantization in the config from Q4_K_M to F16

--- /dev/fd/63	2025-09-20 14:01:13.034904369 +0200
+++ /dev/fd/62	2025-09-20 14:01:13.034904369 +0200
@@ -1,6 +1,6 @@
 {
   "architecture": "amd64",
-  "file_type": "Q4_K_M",
+  "file_type": "F16",
   "model_families": [
     "qwen3moe"
   ],
@@ -8,9 +8,11 @@
   "model_format": "gguf",
   "model_type": "30.5B",
   "os": "linux",
+  "parser": "qwen3-coder",
+  "renderer": "qwen3-coder",
   "rootfs": {
     "diff_ids": [
-      "sha256:1194192cf2a187eb02722edcc3f77b11d21f537048ce04b67ccf8ba78863006a",
+      "sha256:2f3c93d7adf85fcfeb6620d80058b22c51d5a8b21ce18f1c58bd3004c0a63f45",
       "sha256:c6a614465b370a1b4eb95e964f907a24396a5bb842eab6cc730f2cc4c309dc48",
       "sha256:d18a5cc71b84bc4af394a31116bd3932b42241de70c77d2b76d69a314ec8aa12",
       "sha256:69aa441ea44ff5e1e7b56cac4f471e71e8a5e2e3963c29684a9234d5d5e5f7aa"

It doesn't affect the operation of the model but it changes the output from ollama show which may confuse users.

$ diff -u <(ollama show qwen3-coder:30b) <(ollama show qwen3-coder)
--- /dev/fd/63	2025-09-20 14:03:13.136358511 +0200
+++ /dev/fd/62	2025-09-20 14:03:13.136358511 +0200
@@ -3,10 +3,11 @@
     parameters          30.5B       
     context length      262144      
     embedding length    2048        
-    quantization        Q4_K_M      
+    quantization        F16         
 
   Capabilities
     completion    
+    tools         
 
   Parameters
     top_p             0.8

Relevant log output

OS

Linux

GPU

Nvidia

CPU

Intel

Ollama version

0.12.0

Originally created by @rick-github on GitHub (Sep 20, 2025). Original GitHub issue: https://github.com/ollama/ollama/issues/12354 ### What is the issue? The update for tool support has changed the quantization in the config from Q4_K_M to F16 ```diff --- /dev/fd/63 2025-09-20 14:01:13.034904369 +0200 +++ /dev/fd/62 2025-09-20 14:01:13.034904369 +0200 @@ -1,6 +1,6 @@ { "architecture": "amd64", - "file_type": "Q4_K_M", + "file_type": "F16", "model_families": [ "qwen3moe" ], @@ -8,9 +8,11 @@ "model_format": "gguf", "model_type": "30.5B", "os": "linux", + "parser": "qwen3-coder", + "renderer": "qwen3-coder", "rootfs": { "diff_ids": [ - "sha256:1194192cf2a187eb02722edcc3f77b11d21f537048ce04b67ccf8ba78863006a", + "sha256:2f3c93d7adf85fcfeb6620d80058b22c51d5a8b21ce18f1c58bd3004c0a63f45", "sha256:c6a614465b370a1b4eb95e964f907a24396a5bb842eab6cc730f2cc4c309dc48", "sha256:d18a5cc71b84bc4af394a31116bd3932b42241de70c77d2b76d69a314ec8aa12", "sha256:69aa441ea44ff5e1e7b56cac4f471e71e8a5e2e3963c29684a9234d5d5e5f7aa" ``` It doesn't affect the operation of the model but it changes the output from `ollama show` which may confuse users. ```console $ diff -u <(ollama show qwen3-coder:30b) <(ollama show qwen3-coder) --- /dev/fd/63 2025-09-20 14:03:13.136358511 +0200 +++ /dev/fd/62 2025-09-20 14:03:13.136358511 +0200 @@ -3,10 +3,11 @@ parameters 30.5B context length 262144 embedding length 2048 - quantization Q4_K_M + quantization F16 Capabilities completion + tools Parameters top_p 0.8 ``` ### Relevant log output ```shell ``` ### OS Linux ### GPU Nvidia ### CPU Intel ### Ollama version 0.12.0
GiteaMirror added the bug label 2026-04-22 17:09:39 -05:00
GiteaMirror commented 2026-04-22 17:09:41 -05:00
Author
Owner
Copy Link

@rick-github commented on GitHub (Sep 20, 2025):

It seems like the config from qwen3-coder:30b-a3b-fp16 was copied into the other models, since the first diff_id references the image for fp16.

<!-- gh-comment-id:3314944910 --> @rick-github commented on GitHub (Sep 20, 2025): It seems like the config from qwen3-coder:30b-a3b-fp16 was copied into the other models, since the first `diff_id` references the image for fp16.
GiteaMirror commented 2026-04-22 17:09:42 -05:00
Author
Owner
Copy Link

@jmorganca commented on GitHub (Sep 20, 2025):

@rick-github thanks for reporting! This is fixed now – a few (largely inconsequential) metadata fields were out of place temporarily while updating Qwen3-Coder to support tool calling. Sorry about the issue

<!-- gh-comment-id:3314993179 --> @jmorganca commented on GitHub (Sep 20, 2025): @rick-github thanks for reporting! This is fixed now – a few (largely inconsequential) metadata fields were out of place temporarily while updating Qwen3-Coder to support tool calling. Sorry about the issue
Sign in to join this conversation.
No Branch/Tag Specified
Branches Tags
Labels
Clear labels
amd
api
app
bug
build
cli
cloud
compatibility
context-length
create
docker
documentation
embeddings
feature request
feedback wanted
good first issue
gpt-oss
gpu
harmony
help wanted
image
install
intel
js
launch
linux
macos
memory
mlx
model
needs more info
networking
nvidia
ollama.com
performance
pull-request
Mirrored from GitHub Pull Request
 python
question
registry
rendering
thinking
tools
top
vulkan
windows
wsl
No Label bug
Milestone
No items
No Milestone
Projects
Clear projects
No project
Assignees
Clear assignees
GiteaMirror ninjasurge
No Assignees
1 Participants
Notifications
Due Date
No due date set.
Dependencies

No dependencies set.

Reference: github-starred/ollama#33967
Delete Branch "%!s()"

Deleting a branch is permanent. Although the deleted branch may continue to exist for a short time before it actually gets removed, it CANNOT be undone in most cases. Continue?