Coverage for transformer_lens/model_bridge/supported_architectures/__init__.py: 100%
141 statements
« prev ^ index » next coverage.py v7.10.1, created at 2026-09-01 16:23 +0000
« prev ^ index » next coverage.py v7.10.1, created at 2026-09-01 16:23 +0000
1"""Supported architecture adapters.
3This module contains all the supported architecture adapters for different model architectures.
4"""
6from transformer_lens.model_bridge.supported_architectures.afmoe import AfmoeArchitectureAdapter
7from transformer_lens.model_bridge.supported_architectures.apertus import ApertusArchitectureAdapter
8from transformer_lens.model_bridge.supported_architectures.arcee import ArceeArchitectureAdapter
9from transformer_lens.model_bridge.supported_architectures.ast import ASTArchitectureAdapter
10from transformer_lens.model_bridge.supported_architectures.audio_flamingo3 import (
11 AudioFlamingo3ArchitectureAdapter,
12)
13from transformer_lens.model_bridge.supported_architectures.baichuan import (
14 BaichuanArchitectureAdapter,
15)
16from transformer_lens.model_bridge.supported_architectures.bamba import BambaArchitectureAdapter
17from transformer_lens.model_bridge.supported_architectures.bart import BartArchitectureAdapter
18from transformer_lens.model_bridge.supported_architectures.bd3lm import BD3LMArchitectureAdapter
19from transformer_lens.model_bridge.supported_architectures.bert import BertArchitectureAdapter
20from transformer_lens.model_bridge.supported_architectures.bitnet import BitNetArchitectureAdapter
21from transformer_lens.model_bridge.supported_architectures.blenderbot import (
22 BlenderbotArchitectureAdapter,
23)
24from transformer_lens.model_bridge.supported_architectures.bloom import BloomArchitectureAdapter
25from transformer_lens.model_bridge.supported_architectures.codegen import CodeGenArchitectureAdapter
26from transformer_lens.model_bridge.supported_architectures.cohere import (
27 Cohere2ArchitectureAdapter,
28 CohereArchitectureAdapter,
29)
30from transformer_lens.model_bridge.supported_architectures.deepseek_v2 import (
31 DeepSeekV2ArchitectureAdapter,
32)
33from transformer_lens.model_bridge.supported_architectures.deepseek_v3 import (
34 DeepSeekV3ArchitectureAdapter,
35)
36from transformer_lens.model_bridge.supported_architectures.deepseek_v4 import (
37 DeepSeekV4ArchitectureAdapter,
38)
39from transformer_lens.model_bridge.supported_architectures.dream import DreamArchitectureAdapter
40from transformer_lens.model_bridge.supported_architectures.emu3 import Emu3ArchitectureAdapter
41from transformer_lens.model_bridge.supported_architectures.ernie4_5 import (
42 Ernie4_5ArchitectureAdapter,
43)
44from transformer_lens.model_bridge.supported_architectures.ernie4_5_moe import (
45 Ernie4_5_MoeArchitectureAdapter,
46)
47from transformer_lens.model_bridge.supported_architectures.exaone import ExaoneArchitectureAdapter
48from transformer_lens.model_bridge.supported_architectures.exaone4 import Exaone4ArchitectureAdapter
49from transformer_lens.model_bridge.supported_architectures.falcon import FalconArchitectureAdapter
50from transformer_lens.model_bridge.supported_architectures.falcon_h1 import (
51 FalconH1ArchitectureAdapter,
52)
53from transformer_lens.model_bridge.supported_architectures.falcon_mamba import (
54 FalconMambaArchitectureAdapter,
55)
56from transformer_lens.model_bridge.supported_architectures.flex_olmo import (
57 FlexOlmoArchitectureAdapter,
58)
59from transformer_lens.model_bridge.supported_architectures.florence2 import (
60 Florence2ArchitectureAdapter,
61)
62from transformer_lens.model_bridge.supported_architectures.gemma1 import Gemma1ArchitectureAdapter
63from transformer_lens.model_bridge.supported_architectures.gemma2 import Gemma2ArchitectureAdapter
64from transformer_lens.model_bridge.supported_architectures.gemma3 import Gemma3ArchitectureAdapter
65from transformer_lens.model_bridge.supported_architectures.gemma3_multimodal import (
66 Gemma3MultimodalArchitectureAdapter,
67)
68from transformer_lens.model_bridge.supported_architectures.gemma3n import Gemma3nArchitectureAdapter
69from transformer_lens.model_bridge.supported_architectures.gemma4 import Gemma4ArchitectureAdapter
70from transformer_lens.model_bridge.supported_architectures.gemma4_text import (
71 Gemma4TextArchitectureAdapter,
72)
73from transformer_lens.model_bridge.supported_architectures.gidd import GiddArchitectureAdapter
74from transformer_lens.model_bridge.supported_architectures.glm import GlmArchitectureAdapter
75from transformer_lens.model_bridge.supported_architectures.glm4 import Glm4ArchitectureAdapter
76from transformer_lens.model_bridge.supported_architectures.glm4_moe import (
77 Glm4MoeArchitectureAdapter,
78)
79from transformer_lens.model_bridge.supported_architectures.glm4_moe_lite import (
80 Glm4MoeLiteArchitectureAdapter,
81)
82from transformer_lens.model_bridge.supported_architectures.glm4v import Glm4vArchitectureAdapter
83from transformer_lens.model_bridge.supported_architectures.glm_asr import GlmAsrArchitectureAdapter
84from transformer_lens.model_bridge.supported_architectures.glm_moe_dsa import (
85 GlmMoeDsaArchitectureAdapter,
86)
87from transformer_lens.model_bridge.supported_architectures.gpt2 import GPT2ArchitectureAdapter
88from transformer_lens.model_bridge.supported_architectures.gpt2_lm_head_custom import (
89 Gpt2LmHeadCustomArchitectureAdapter,
90)
91from transformer_lens.model_bridge.supported_architectures.gpt_bigcode import (
92 GPTBigCodeArchitectureAdapter,
93)
94from transformer_lens.model_bridge.supported_architectures.gpt_oss import GPTOSSArchitectureAdapter
95from transformer_lens.model_bridge.supported_architectures.gptj import GptjArchitectureAdapter
96from transformer_lens.model_bridge.supported_architectures.granite import GraniteArchitectureAdapter
97from transformer_lens.model_bridge.supported_architectures.granite_moe import (
98 GraniteMoeArchitectureAdapter,
99)
100from transformer_lens.model_bridge.supported_architectures.granite_moe_hybrid import (
101 GraniteMoeHybridArchitectureAdapter,
102)
103from transformer_lens.model_bridge.supported_architectures.hrm_text import (
104 HrmTextArchitectureAdapter,
105)
106from transformer_lens.model_bridge.supported_architectures.hubert import HubertArchitectureAdapter
107from transformer_lens.model_bridge.supported_architectures.hunyuan_v1_dense import (
108 HunYuanDenseV1ArchitectureAdapter,
109)
110from transformer_lens.model_bridge.supported_architectures.hyenadna import (
111 HyenaDNAArchitectureAdapter,
112)
113from transformer_lens.model_bridge.supported_architectures.idefics3 import (
114 Idefics3ArchitectureAdapter,
115)
116from transformer_lens.model_bridge.supported_architectures.internlm2 import (
117 InternLM2ArchitectureAdapter,
118)
119from transformer_lens.model_bridge.supported_architectures.jais2 import Jais2ArchitectureAdapter
120from transformer_lens.model_bridge.supported_architectures.jamba import JambaArchitectureAdapter
121from transformer_lens.model_bridge.supported_architectures.jetmoe import JetMoeArchitectureAdapter
122from transformer_lens.model_bridge.supported_architectures.laguna import LagunaArchitectureAdapter
123from transformer_lens.model_bridge.supported_architectures.led import LEDArchitectureAdapter
124from transformer_lens.model_bridge.supported_architectures.lfm2 import Lfm2ArchitectureAdapter
125from transformer_lens.model_bridge.supported_architectures.lfm2_moe import (
126 Lfm2MoeArchitectureAdapter,
127)
128from transformer_lens.model_bridge.supported_architectures.llada import LLaDAArchitectureAdapter
129from transformer_lens.model_bridge.supported_architectures.llada2_moe import (
130 LLaDA2MoeArchitectureAdapter,
131)
132from transformer_lens.model_bridge.supported_architectures.llama import LlamaArchitectureAdapter
133from transformer_lens.model_bridge.supported_architectures.llama4 import Llama4ArchitectureAdapter
134from transformer_lens.model_bridge.supported_architectures.llama4_multimodal import (
135 Llama4MultimodalArchitectureAdapter,
136)
137from transformer_lens.model_bridge.supported_architectures.llava import LlavaArchitectureAdapter
138from transformer_lens.model_bridge.supported_architectures.llava_next import (
139 LlavaNextArchitectureAdapter,
140)
141from transformer_lens.model_bridge.supported_architectures.llava_onevision import (
142 LlavaOnevisionArchitectureAdapter,
143)
144from transformer_lens.model_bridge.supported_architectures.longt5 import LongT5ArchitectureAdapter
145from transformer_lens.model_bridge.supported_architectures.m2m100 import M2M100ArchitectureAdapter
146from transformer_lens.model_bridge.supported_architectures.mamba import MambaArchitectureAdapter
147from transformer_lens.model_bridge.supported_architectures.mamba2 import Mamba2ArchitectureAdapter
148from transformer_lens.model_bridge.supported_architectures.marian import MarianArchitectureAdapter
149from transformer_lens.model_bridge.supported_architectures.mbart import MBartArchitectureAdapter
150from transformer_lens.model_bridge.supported_architectures.mingpt import MingptArchitectureAdapter
151from transformer_lens.model_bridge.supported_architectures.minimax_m2 import (
152 MiniMaxM2ArchitectureAdapter,
153)
154from transformer_lens.model_bridge.supported_architectures.ministral3 import (
155 Ministral3ArchitectureAdapter,
156)
157from transformer_lens.model_bridge.supported_architectures.mistral import MistralArchitectureAdapter
158from transformer_lens.model_bridge.supported_architectures.mistral3 import (
159 Mistral3ArchitectureAdapter,
160)
161from transformer_lens.model_bridge.supported_architectures.mixtral import MixtralArchitectureAdapter
162from transformer_lens.model_bridge.supported_architectures.modernbert_decoder import (
163 ModernBertDecoderArchitectureAdapter,
164)
165from transformer_lens.model_bridge.supported_architectures.mpt import MPTArchitectureAdapter
166from transformer_lens.model_bridge.supported_architectures.music_flamingo import (
167 MusicFlamingoArchitectureAdapter,
168)
169from transformer_lens.model_bridge.supported_architectures.nanochat import (
170 NanoChatArchitectureAdapter,
171)
172from transformer_lens.model_bridge.supported_architectures.nanogpt import NanogptArchitectureAdapter
173from transformer_lens.model_bridge.supported_architectures.native import NativeArchitectureAdapter
174from transformer_lens.model_bridge.supported_architectures.neel_solu_old import (
175 NeelSoluOldArchitectureAdapter,
176)
177from transformer_lens.model_bridge.supported_architectures.nemotron import (
178 NemotronArchitectureAdapter,
179)
180from transformer_lens.model_bridge.supported_architectures.nemotron_h import (
181 NemotronHArchitectureAdapter,
182)
183from transformer_lens.model_bridge.supported_architectures.neo import NeoArchitectureAdapter
184from transformer_lens.model_bridge.supported_architectures.neox import NeoxArchitectureAdapter
185from transformer_lens.model_bridge.supported_architectures.olmo import OlmoArchitectureAdapter
186from transformer_lens.model_bridge.supported_architectures.olmo2 import Olmo2ArchitectureAdapter
187from transformer_lens.model_bridge.supported_architectures.olmo3 import Olmo3ArchitectureAdapter
188from transformer_lens.model_bridge.supported_architectures.olmo_hybrid import (
189 OlmoHybridArchitectureAdapter,
190)
191from transformer_lens.model_bridge.supported_architectures.olmoe import OlmoeArchitectureAdapter
192from transformer_lens.model_bridge.supported_architectures.openai_gpt import (
193 OpenAIGPTArchitectureAdapter,
194)
195from transformer_lens.model_bridge.supported_architectures.openelm import OpenElmArchitectureAdapter
196from transformer_lens.model_bridge.supported_architectures.opt import OptArchitectureAdapter
197from transformer_lens.model_bridge.supported_architectures.ouro import OuroArchitectureAdapter
198from transformer_lens.model_bridge.supported_architectures.pegasus import PegasusArchitectureAdapter
199from transformer_lens.model_bridge.supported_architectures.phi import PhiArchitectureAdapter
200from transformer_lens.model_bridge.supported_architectures.phi3 import Phi3ArchitectureAdapter
201from transformer_lens.model_bridge.supported_architectures.phimoe import PhiMoEArchitectureAdapter
202from transformer_lens.model_bridge.supported_architectures.pretrain import (
203 PretrainArchitectureAdapter,
204)
205from transformer_lens.model_bridge.supported_architectures.qwen import QwenArchitectureAdapter
206from transformer_lens.model_bridge.supported_architectures.qwen2 import Qwen2ArchitectureAdapter
207from transformer_lens.model_bridge.supported_architectures.qwen2_5_vl import (
208 Qwen2_5_VLArchitectureAdapter,
209)
210from transformer_lens.model_bridge.supported_architectures.qwen2_audio import (
211 Qwen2AudioArchitectureAdapter,
212)
213from transformer_lens.model_bridge.supported_architectures.qwen2_moe import (
214 Qwen2MoeArchitectureAdapter,
215)
216from transformer_lens.model_bridge.supported_architectures.qwen3 import Qwen3ArchitectureAdapter
217from transformer_lens.model_bridge.supported_architectures.qwen3_5 import Qwen3_5ArchitectureAdapter
218from transformer_lens.model_bridge.supported_architectures.qwen3_5_moe import (
219 Qwen3_5MoeArchitectureAdapter,
220 Qwen3_5MoeMultimodalArchitectureAdapter,
221)
222from transformer_lens.model_bridge.supported_architectures.qwen3_5_multimodal import (
223 Qwen3_5MultimodalArchitectureAdapter,
224)
225from transformer_lens.model_bridge.supported_architectures.qwen3_moe import (
226 Qwen3MoeArchitectureAdapter,
227)
228from transformer_lens.model_bridge.supported_architectures.qwen3_next import (
229 Qwen3NextArchitectureAdapter,
230)
231from transformer_lens.model_bridge.supported_architectures.qwen3_vl import (
232 Qwen3VLArchitectureAdapter,
233)
234from transformer_lens.model_bridge.supported_architectures.qwen3_vl_moe import (
235 Qwen3VLMoeArchitectureAdapter,
236)
237from transformer_lens.model_bridge.supported_architectures.raven import RavenArchitectureAdapter
238from transformer_lens.model_bridge.supported_architectures.recurrent_gemma import (
239 RecurrentGemmaArchitectureAdapter,
240)
241from transformer_lens.model_bridge.supported_architectures.rwkv import RwkvArchitectureAdapter
242from transformer_lens.model_bridge.supported_architectures.rwkv7 import RWKV7ArchitectureAdapter
243from transformer_lens.model_bridge.supported_architectures.seed_oss import (
244 SeedOssArchitectureAdapter,
245)
246from transformer_lens.model_bridge.supported_architectures.smollm3 import SmolLM3ArchitectureAdapter
247from transformer_lens.model_bridge.supported_architectures.stablelm import (
248 StableLmArchitectureAdapter,
249)
250from transformer_lens.model_bridge.supported_architectures.starcoder2 import (
251 Starcoder2ArchitectureAdapter,
252)
253from transformer_lens.model_bridge.supported_architectures.switch_transformers import (
254 SwitchTransformersArchitectureAdapter,
255)
256from transformer_lens.model_bridge.supported_architectures.vit import (
257 ViTArchitectureAdapter,
258)
259from transformer_lens.model_bridge.supported_architectures.t5 import T5ArchitectureAdapter
260from transformer_lens.model_bridge.supported_architectures.t5gemma import T5GemmaArchitectureAdapter
261from transformer_lens.model_bridge.supported_architectures.t5gemma2 import (
262 T5Gemma2ArchitectureAdapter,
263)
264from transformer_lens.model_bridge.supported_architectures.vaultgemma import (
265 VaultGemmaArchitectureAdapter,
266)
267from transformer_lens.model_bridge.supported_architectures.xglm import XGLMArchitectureAdapter
268from transformer_lens.model_bridge.supported_architectures.youtu import YoutuArchitectureAdapter
269from transformer_lens.model_bridge.supported_architectures.zamba2 import Zamba2ArchitectureAdapter
271__all__ = [
272 "AfmoeArchitectureAdapter",
273 "ApertusArchitectureAdapter",
274 "ArceeArchitectureAdapter",
275 "ASTArchitectureAdapter",
276 "AudioFlamingo3ArchitectureAdapter",
277 "BD3LMArchitectureAdapter",
278 "BaichuanArchitectureAdapter",
279 "BambaArchitectureAdapter",
280 "BartArchitectureAdapter",
281 "BertArchitectureAdapter",
282 "BitNetArchitectureAdapter",
283 "BlenderbotArchitectureAdapter",
284 "BloomArchitectureAdapter",
285 "CodeGenArchitectureAdapter",
286 "Cohere2ArchitectureAdapter",
287 "CohereArchitectureAdapter",
288 "DeepSeekV2ArchitectureAdapter",
289 "DeepSeekV3ArchitectureAdapter",
290 "DreamArchitectureAdapter",
291 "Emu3ArchitectureAdapter",
292 "Ernie4_5ArchitectureAdapter",
293 "Ernie4_5_MoeArchitectureAdapter",
294 "ExaoneArchitectureAdapter",
295 "Exaone4ArchitectureAdapter",
296 "FalconArchitectureAdapter",
297 "FalconH1ArchitectureAdapter",
298 "FalconMambaArchitectureAdapter",
299 "FlexOlmoArchitectureAdapter",
300 "Florence2ArchitectureAdapter",
301 "Gemma1ArchitectureAdapter",
302 "Gemma2ArchitectureAdapter",
303 "Gemma3ArchitectureAdapter",
304 "Gemma3nArchitectureAdapter",
305 "Gemma3MultimodalArchitectureAdapter",
306 "Gemma4ArchitectureAdapter",
307 "Gemma4TextArchitectureAdapter",
308 "GiddArchitectureAdapter",
309 "GlmArchitectureAdapter",
310 "Glm4ArchitectureAdapter",
311 "Glm4vArchitectureAdapter",
312 "GlmAsrArchitectureAdapter",
313 "GlmMoeDsaArchitectureAdapter",
314 "Glm4MoeArchitectureAdapter",
315 "Glm4MoeLiteArchitectureAdapter",
316 "GraniteArchitectureAdapter",
317 "GraniteMoeArchitectureAdapter",
318 "GraniteMoeHybridArchitectureAdapter",
319 "GPT2ArchitectureAdapter",
320 "GPTBigCodeArchitectureAdapter",
321 "GPTOSSArchitectureAdapter",
322 "Gpt2LmHeadCustomArchitectureAdapter",
323 "GptjArchitectureAdapter",
324 "HubertArchitectureAdapter",
325 "HyenaDNAArchitectureAdapter",
326 "HunYuanDenseV1ArchitectureAdapter",
327 "Idefics3ArchitectureAdapter",
328 "InternLM2ArchitectureAdapter",
329 "LEDArchitectureAdapter",
330 "LagunaArchitectureAdapter",
331 "Jais2ArchitectureAdapter",
332 "JambaArchitectureAdapter",
333 "JetMoeArchitectureAdapter",
334 "Ministral3ArchitectureAdapter",
335 "VaultGemmaArchitectureAdapter",
336 "YoutuArchitectureAdapter",
337 "LLaDA2MoeArchitectureAdapter",
338 "LlamaArchitectureAdapter",
339 "Llama4ArchitectureAdapter",
340 "Llama4MultimodalArchitectureAdapter",
341 "LlavaArchitectureAdapter",
342 "LlavaNextArchitectureAdapter",
343 "LlavaOnevisionArchitectureAdapter",
344 "Lfm2MoeArchitectureAdapter",
345 "LongT5ArchitectureAdapter",
346 "M2M100ArchitectureAdapter",
347 "MambaArchitectureAdapter",
348 "Mamba2ArchitectureAdapter",
349 "MarianArchitectureAdapter",
350 "MBartArchitectureAdapter",
351 "MingptArchitectureAdapter",
352 "MiniMaxM2ArchitectureAdapter",
353 "MistralArchitectureAdapter",
354 "Mistral3ArchitectureAdapter",
355 "MixtralArchitectureAdapter",
356 "ModernBertDecoderArchitectureAdapter",
357 "MPTArchitectureAdapter",
358 "MusicFlamingoArchitectureAdapter",
359 "NanogptArchitectureAdapter",
360 "NanoChatArchitectureAdapter",
361 "NativeArchitectureAdapter",
362 "NeelSoluOldArchitectureAdapter",
363 "NemotronArchitectureAdapter",
364 "NemotronHArchitectureAdapter",
365 "NeoArchitectureAdapter",
366 "NeoxArchitectureAdapter",
367 "OpenAIGPTArchitectureAdapter",
368 "OpenElmArchitectureAdapter",
369 "PegasusArchitectureAdapter",
370 "OlmoArchitectureAdapter",
371 "Olmo2ArchitectureAdapter",
372 "Olmo3ArchitectureAdapter",
373 "OlmoHybridArchitectureAdapter",
374 "OlmoeArchitectureAdapter",
375 "OptArchitectureAdapter",
376 "PhiArchitectureAdapter",
377 "Phi3ArchitectureAdapter",
378 "PhiMoEArchitectureAdapter",
379 "PretrainArchitectureAdapter",
380 "QwenArchitectureAdapter",
381 "Qwen2ArchitectureAdapter",
382 "Qwen2_5_VLArchitectureAdapter",
383 "Qwen2AudioArchitectureAdapter",
384 "Qwen2MoeArchitectureAdapter",
385 "Qwen3ArchitectureAdapter",
386 "Qwen3VLArchitectureAdapter",
387 "Qwen3VLMoeArchitectureAdapter",
388 "RavenArchitectureAdapter",
389 "RwkvArchitectureAdapter",
390 "Qwen3MoeArchitectureAdapter",
391 "Qwen3NextArchitectureAdapter",
392 "Qwen3_5ArchitectureAdapter",
393 "Qwen3_5MoeArchitectureAdapter",
394 "Qwen3_5MoeMultimodalArchitectureAdapter",
395 "Qwen3_5MultimodalArchitectureAdapter",
396 "RecurrentGemmaArchitectureAdapter",
397 "SeedOssArchitectureAdapter",
398 "SmolLM3ArchitectureAdapter",
399 "StableLmArchitectureAdapter",
400 "SwitchTransformersArchitectureAdapter",
401 "Starcoder2ArchitectureAdapter",
402 "T5ArchitectureAdapter",
403 "T5GemmaArchitectureAdapter",
404 "T5Gemma2ArchitectureAdapter",
405 "ViTArchitectureAdapter",
406 "XGLMArchitectureAdapter",
407 "Zamba2ArchitectureAdapter",
408 "HrmTextArchitectureAdapter",
409 "LLaDAArchitectureAdapter",
410 "Lfm2ArchitectureAdapter",
411 "OuroArchitectureAdapter",
412 "RWKV7ArchitectureAdapter",
413 "T5Gemma2ArchitectureAdapter",
414]