From f33a567383560533e4b9a78cd32837df576ae98d Mon Sep 17 00:00:00 2001 From: metcalfc Date: Sun, 27 Jul 2025 02:52:24 +0000 Subject: [PATCH] Update Together.ai model configurations --- .version_cache.json | 18 ++++++++++++++++++ ...235b-a22b-instruct-2507-fp8-throughput.yaml | 15 +++++++++++++++ blocks/public/qwen3-32b-fp8.yaml | 15 +++++++++++++++ .../qwen3-coder-480b-a35b-instruct-fp8.yaml | 15 +++++++++++++++ 4 files changed, 63 insertions(+) create mode 100644 blocks/public/qwen3-235b-a22b-instruct-2507-fp8-throughput.yaml create mode 100644 blocks/public/qwen3-32b-fp8.yaml create mode 100644 blocks/public/qwen3-coder-480b-a35b-instruct-fp8.yaml diff --git a/.version_cache.json b/.version_cache.json index a41e5d8..7871f14 100644 --- a/.version_cache.json +++ b/.version_cache.json @@ -538,5 +538,23 @@ "version": "1.1.0", "filename": "kimi-k2-instruct.yaml", "display_name": "Kimi K2 Instruct" + }, + "Qwen/Qwen3-Coder-480B-A35B-Instruct-FP8": { + "hash": "e20b2d129744ccffba827a363c2851d9", + "version": "1.1.0", + "filename": "qwen3-coder-480b-a35b-instruct-fp8.yaml", + "display_name": "Qwen3 Coder 480B A35B Instruct Fp8" + }, + "Qwen/Qwen3-32B-FP8": { + "hash": "a7d6ace616c4f3586f25dba6d1ba13f7", + "version": "1.1.0", + "filename": "qwen3-32b-fp8.yaml", + "display_name": "Qwen3 32B Fp8" + }, + "Qwen/Qwen3-235B-A22B-Instruct-2507-tput": { + "hash": "00022257737c36e190a2311f8ea6f771", + "version": "1.1.0", + "filename": "qwen3-235b-a22b-instruct-2507-fp8-throughput.yaml", + "display_name": "Qwen3 235B A22B Instruct 2507 FP8 Throughput" } } \ No newline at end of file diff --git a/blocks/public/qwen3-235b-a22b-instruct-2507-fp8-throughput.yaml b/blocks/public/qwen3-235b-a22b-instruct-2507-fp8-throughput.yaml new file mode 100644 index 0000000..d5e1507 --- /dev/null +++ b/blocks/public/qwen3-235b-a22b-instruct-2507-fp8-throughput.yaml @@ -0,0 +1,15 @@ +--- +name: Qwen3 235B A22B Instruct 2507 FP8 Throughput +version: 1.1.0 +schema: v1 +models: + - name: Qwen3 235B A22B Instruct 2507 FP8 Throughput + provider: together + model: Qwen/Qwen3-235B-A22B-Instruct-2507-tput + apiKey: ${{ inputs.TOGETHER_API_KEY }} + defaultCompletionOptions: + contextLength: 262144 + roles: + - apply + - chat + - edit diff --git a/blocks/public/qwen3-32b-fp8.yaml b/blocks/public/qwen3-32b-fp8.yaml new file mode 100644 index 0000000..20faacc --- /dev/null +++ b/blocks/public/qwen3-32b-fp8.yaml @@ -0,0 +1,15 @@ +--- +name: Qwen3 32B Fp8 +version: 1.1.0 +schema: v1 +models: + - name: Qwen3 32B Fp8 + provider: together + model: Qwen/Qwen3-32B-FP8 + apiKey: ${{ inputs.TOGETHER_API_KEY }} + defaultCompletionOptions: + contextLength: 40960 + roles: + - apply + - chat + - edit diff --git a/blocks/public/qwen3-coder-480b-a35b-instruct-fp8.yaml b/blocks/public/qwen3-coder-480b-a35b-instruct-fp8.yaml new file mode 100644 index 0000000..d687d06 --- /dev/null +++ b/blocks/public/qwen3-coder-480b-a35b-instruct-fp8.yaml @@ -0,0 +1,15 @@ +--- +name: Qwen3 Coder 480B A35B Instruct Fp8 +version: 1.1.0 +schema: v1 +models: + - name: Qwen3 Coder 480B A35B Instruct Fp8 + provider: together + model: Qwen/Qwen3-Coder-480B-A35B-Instruct-FP8 + apiKey: ${{ inputs.TOGETHER_API_KEY }} + defaultCompletionOptions: + contextLength: 262144 + roles: + - apply + - chat + - edit