Added Qwen2.5-7b-coder with the updated weights

The Qwen team still calls it Qwen2.5, but as can be seen from the
benchmarks the difference in performance compared to the old weights
is pretty substantial. The GGUF version of this model made by Bartowski
calls it 2.5.1 to differentiate it from the earlier version of the
same model.
This commit is contained in:
Jaap Buurman 2024-11-07 13:18:24 +01:00
parent c84f2996ec
commit af0466ea83

View file

@ -1634,3 +1634,26 @@
versions: 0.61.1.dev
seconds_per_case: 18.4
total_cost: 0.0000
- dirname: 2024-11-07-06-15-36--Qwen2.5.1-Coder-7B-Instruct-GGUF:Q8_0-32k-whole
test_cases: 133
model: ollama/Qwen2.5.1-Coder-7B-Instruct-GGUF:Q8_0-32k
edit_format: whole
commit_hash: e76704e
pass_rate_1: 52.6
pass_rate_2: 63.9
percent_cases_well_formed: 100.0
error_outputs: 0
num_malformed_responses: 0
num_with_malformed_responses: 0
user_asks: 4
lazy_comments: 0
syntax_errors: 0
indentation_errors: 0
exhausted_context_windows: 0
test_timeouts: 1
command: aider --model ollama/Qwen2.5.1-Coder-7B-Instruct-GGUF:Q8_0-32k
date: 2024-11-07
versions: 0.59.2.dev
seconds_per_case: 18.2
total_cost: 0.0000