Add v1.3 release notes.

[joshuayao]

Add v1.3 release notes.

[Copilot]

Copilot reviewed 2 out of 2 changed files in this pull request and generated 1 comment.

Comments suppressed due to low confidence (2)

release_notes/v1.3.md:2

• The TODO placeholder for the number of pull requests may confuse users. Consider updating this placeholder with the final count before release.

We are excited to announce the release of OPEA version 1.3, which includes significant contributions from the open-source community. This release addresses over 400(TODO: the number is subject to change) pull requests.

release_notes/RELEASE.md:18

• [nitpick] The date format seems inconsistent with other entries (e.g., using 'Apr' instead of 'July'). Consider standardizing abbreviations to maintain consistency in the release schedule.

| 1.4     | July 2025    |

[mkbhanda]

Does AMD want to add to the table of supported models on serving frameworks and hardware?
Arc not mentioned in table, do we want to add?

[Yu-amd]

AMD team has a validated model matrix we can provide.

[Yu-amd]

AMD team has a validated model matrix we can provide.
| Model | TGI-Gaudi | vLLM-CPU | vLLM-Gaudi | vLLM-ROCm | OVMS | Optimum-Habana | PredictionGuard |
| ------------------------------------------- | --------- | -------- | ---------- | --------- | -------- | -------------- | --------------- |
| deepseek-ai/DeepSeek-R1-Distill-Llama-8B | ✓ | - | ✓ | ✓ | - | ✓ | - |
| deepseek-ai/DeepSeek-R1-Distill-Llama-70B | ✓ | - | ✓ | ✓ | - | ✓ | - |
| deepseek-ai/DeepSeek-R1-Distill-Qwen-1.5B | ✓ | - | ✓ | ✓ | - | ✓ | - |
| deepseek-ai/DeepSeek-R1-Distill-Qwen-7B | ✓ | - | ✓ | ✓ | - | ✓ | - |
| deepseek-ai/DeepSeek-R1-Distill-Qwen-14B | ✓ | - | ✓ | ✓ | - | ✓ | - |
| deepseek-ai/DeepSeek-R1-Distill-Qwen-32B | ✓ | - | ✓ | ✓ | - | ✓ | - |
| deepseek-ai/Deepseek-v3 | ✓ | - | ✓ | ✓ | - | ✓ | - |
| Hermes-3-Llama-3.1-8B | - | - | - | ✓ | - | - | ✓ |
| ibm-granite/granite-3.2-8b-instruct | - | - | ✓ | ✓ | - | | - |
| Phi-4-mini | x | x | x | ✓ | x | ✓ | - | | Phi-4-multimodal-instruct | x | x | x | ✓ | x | ✓ | - |
| mistralai/Mistral-Small-24B-Instruct-2501 | ✓ | - | ✓ | ✓ | - | ✓ | - |
| mistralai/Mistral-Large-Instruct-2411 | x | - | ✓ | ✓ | - | ✓ | -
|
Phi-4-mini works with latest docker image: rocm/vllm:rocm6.3.1_instinct_vllm0.8.3_20250410

@joshuayao Please see the table above, which includes a column of validated models on ROCm.

Thanks!

[ashahba]

LGTM!

[joshuayao]

AMD team has a validated model matrix we can provide.

Model	TGI-Gaudi	vLLM-CPU	vLLM-Gaudi	vLLM-ROCm	OVMS	Optimum-Habana	PredictionGuard
deepseek-ai/DeepSeek-R1-Distill-Llama-8B	✓	-	✓	✓	-	✓	-
deepseek-ai/DeepSeek-R1-Distill-Llama-70B	✓	-	✓	✓	-	✓	-
deepseek-ai/DeepSeek-R1-Distill-Qwen-1.5B	✓	-	✓	✓	-	✓	-
deepseek-ai/DeepSeek-R1-Distill-Qwen-7B	✓	-	✓	✓	-	✓	-
deepseek-ai/DeepSeek-R1-Distill-Qwen-14B	✓	-	✓	✓	-	✓	-
deepseek-ai/DeepSeek-R1-Distill-Qwen-32B	✓	-	✓	✓	-	✓	-
deepseek-ai/Deepseek-v3	✓	-	✓	✓	-	✓	-
Hermes-3-Llama-3.1-8B	-	-	-	✓	-	-	✓
ibm-granite/granite-3.2-8b-instruct	-	-	✓	✓	-		-
Phi-4-mini	x	x	x	✓	x	✓	-
mistralai/Mistral-Small-24B-Instruct-2501	✓	-	✓	✓	-	✓	-
mistralai/Mistral-Large-Instruct-2411	x	-	✓	✓	-	✓	-

|
Phi-4-mini works with latest docker image: rocm/vllm:rocm6.3.1_instinct_vllm0.8.3_20250410

@joshuayao Please see the table above, which includes a column of validated models on ROCm.

Thanks!

Hi @Yu-amd Updated. Thanks for your information.

[Yu-amd]

LGTM. Thank you.

[eero-t]

Sorry, I had typos in my suggestions.

[joshuayao]

Sorry, I had typos in my suggestions.

No problem. Thanks for the updates.

[eero-t]

I guess "Other Notable Changes" section content came directly from the PR titles, but for the release notes I think it would good to use correct & consistent capitalization for the names (and maybe also monotype for function names).

Following terms seemed to have instances with incorrect capitalization:

• AgentQnA
• Dataprep
• DeepSeek
• Docker
• Helm
• LlamaGuard
• LVM
• Milvus
• MySQL
• OpenAI
• PubMed
• Python
• RAG
• TEI
• vLLM
• Xeon
• YAML