[bong-water-water-bong]

429044 text-generation models on HF, 340379 with an architecture (2894 architectures).

	models	share
mapped (a backend's code accepts the architecture)	321755	94.53%
checked (passed serve_e2e on Strix Halo)	217141	63.79%

backend	mapped	checked
hrx	94.53%	63.79%
npu	53.7%	9.41%

Unmapped architectures with the most models:

• Step1MoEForCausalLM: 2882
• ParlerTTSForConditionalGeneration: 1585
• ChessForCausalLM: 970
• LlavaLlamaForCausalLM: 453
• DynamicAlibiForCausalLM: 245
• OpenLMForCausalLM: 245
• DynamicForgettingForCausalLM: 240
• BertLMHeadModel: 212
• MobileLLMForCausalLM: 176
• BartForConditionalGeneration: 174

[github-actions]

PR Reviewer Guide 🔍

Here are some key observations to aid the review process:

⏱️ Estimated effort to review: 2 🔵🔵⚪⚪⚪

🧪 No relevant tests

🔒 No security concerns identified

⚡ Recommended focus areas for review Data Consistency The census data shows an increase in the number of models with architectures (340379) and mapped models (321755) from 428400 total models, but the no_arch count increased from 88380 to 88665, which is inconsistent with the total model count increase. This discrepancy suggests that either the total count or the no_arch count may be incorrect, or there's a data collection issue that needs to be verified. "total": 429044, "no_arch": 88665, "pages": 430, Missing Documentation The PR description includes a table of backend coverage but does not specify how the numbers were measured on the hardware. The PR body should include measurements for each number mentioned in the description to ensure claims are substantiated. {