sleepy/llama.cpp
1
0
Fork 0
Code Issues 8 Pull Requests 1 Actions 175 Packages Projects Releases Wiki Activity

model : Qwen3 Next (#16095)

Browse Source 
* Qwen3 Next - cleaned up version

* Whitespaces and stuff

* Correct minor errors

* Update src/llama-model.cpp

Co-authored-by: Sigbjørn Skjæret <sigbjorn.skjaeret@scala.com>

* Misc. fixes.

* Clean up code, add missing hybrid qualifier

* Did someone transpose the SOLVE_TRI result matrix? Perhaps...

* Whitespace

* Proper tensors for cb calls

* Use llama-graph.h vertical alignment

* BROKEN: chunking

* Set new tensors as inputs.

* Proper chunk logic

* It's the circle of life...

* More shenanigans for n_seq > 1

* Nail in the coffin?

* Fix Windows build

* Eh, one fails on Windows, the other fails on Mac... just use general capture.

* quant : cleanup

* model : cleanup

* qwen3 : cleanup

* cont : cleanup

* cont : cleanup

* ggml : revert change

* qwen3 : cleanup

* cont : cleanup

* Readd cmath

* qwen3 : fix typo

* Update convert_hf_to_gguf.py

Co-authored-by: Sigbjørn Skjæret <sigbjorn.skjaeret@scala.com>

* Usual suspects

* fix my bad suggestion

---------

Co-authored-by: Sigbjørn Skjæret <sigbjorn.skjaeret@scala.com>
Co-authored-by: Georgi Gerganov <ggerganov@gmail.com>
This commit is contained in:
Piotr Wilkin (ilintar)
2025-11-28 12:02:56 +01:00
committed by GitHub
parent 73955f7d2a
commit ff55414c42
16 changed files with 1345 additions and 19 deletions
Download Patch File Download Diff File Expand all files Collapse all files
examples/model-conversion/scripts/causal/run-converted-model.sh
+7 -1
View File
@@ -4,6 +4,11 @@ set -e
# First try command line argument, then environment variable, then file
CONVERTED_MODEL="${1:-"$CONVERTED_MODEL"}"
MODEL_TESTING_PROMPT="${2:-"$MODEL_TESTING_PROMPT"}"
if [ -z "$MODEL_TESTING_PROMPT"]; then
MODEL_TESTING_PROMPT="Hello, my name is"
fi
# Final check if we have a model path
if [ -z "$CONVERTED_MODEL" ]; then
@@ -14,7 +19,8 @@ if [ -z "$CONVERTED_MODEL" ]; then
fi
echo $CONVERTED_MODEL
echo $MODEL_TESTING_PROMPT
cmake --build ../../build --target llama-logits -j8
../../build/bin/llama-logits -m "$CONVERTED_MODEL" "Hello, my name is"
../../build/bin/llama-logits -m "$CONVERTED_MODEL" "$MODEL_TESTING_PROMPT"
examples/model-conversion/scripts/causal/run-org-model.py
+6 -2
View File
@@ -184,8 +184,12 @@ model_name = os.path.basename(model_path)
# of using AutoModelForCausalLM.
print(f"Model class: {model.__class__.__name__}")
prompt = "Hello, my name is"
input_ids = tokenizer(prompt, return_tensors="pt").input_ids
device = next(model.parameters()).device
if os.getenv("MODEL_TESTING_PROMPT"):
prompt = os.getenv("MODEL_TESTING_PROMPT")
else:
prompt = "Hello, my name is"
input_ids = tokenizer(prompt, return_tensors="pt").input_ids.to(device)
print(f"Input tokens: {input_ids}")
print(f"Input text: {repr(prompt)}")