Models / dots_ocr

dots.mocr

Reconstructed from its own config.json with no weights read. 599K downloads on Hugging Face.

Our count against the checkpoint

The left number comes from the graph. The right one is the number of scalars in the published weight files. Nothing on this page was tuned to make them agree.

Derived from structure

2.72B

2,722,830,976 parameters

In the published checkpoint

3.04B

3,039,179,264 scalars · safetensors.total, read 2026-07-04

Delta

-10.4%

custom-code This repository ships its own modeling code (`auto_map`, e.g. `configuration_dots.py`), so `config.json` names a class in the repo rather than an architecture `transformers` defines. The graph below is what those config keys mean under `transformers` semantics, which is not necessarily what the repo's own file builds. A gap here is a statement about what we read, not about the checkpoint.

What it costs to run

Cost is a roofline estimate on the priced GPU for 10 epochs at batch 32 over 50,000 samples (assumed; no dataset attached). GPU fit is fp32 weights plus gradients plus two Adam moments (16 bytes per parameter) with 1.3x headroom; activations are not included and grow with batch size.

Layers

345

Will it forward-pass

Yes

Priced on

A10G (24GB)

Est. one run

$35360.12

Card	Memory
T4 (16GB)	weights + activations	does not fit
A100 (40GB)	weights + activations	does not fit
H100 (80GB)	weights + activations	fits

Structure

348 nodes. Output shapes are propagated from the input shape, batch dimension excluded.

	Layer	Type	Output shape
1	Input	Input	1 × 131072
2	Embedding	Embedding	1 × 131072 × 1536
3	RoPE	RoPE	1 × 131072 × 1536
4	Vision input	Input	3 × 224 × 224
5	PatchEmbed	Patch Embed	256 × 1536
6	Patch_Position_Embedding	Learned Pos Embed	256 × 1536
7	Vision_LN_1	LayerNorm	256 × 1536
8	Vision_Attn_1	Multi-Head Attention	256 × 1536
9	Vision_Add_1	Add	256 × 1536
10	Vision_FFN_1	Feed Forward	256 × 1536
11	Vision_LN_2	LayerNorm	256 × 1536
12	Vision_Attn_2	Multi-Head Attention	256 × 1536
13	Vision_Add_2	Add	256 × 1536
14	Vision_FFN_2	Feed Forward	256 × 1536
15	Vision_LN_3	LayerNorm	256 × 1536
16	Vision_Attn_3	Multi-Head Attention	256 × 1536
17	Vision_Add_3	Add	256 × 1536
18	Vision_FFN_3	Feed Forward	256 × 1536
19	Vision_LN_4	LayerNorm	256 × 1536
20	Vision_Attn_4	Multi-Head Attention	256 × 1536
21	Vision_Add_4	Add	256 × 1536
22	Vision_FFN_4	Feed Forward	256 × 1536
23	Vision_LN_5	LayerNorm	256 × 1536
24	Vision_Attn_5	Multi-Head Attention	256 × 1536
25	Vision_Add_5	Add	256 × 1536
26	Vision_FFN_5	Feed Forward	256 × 1536
27	Vision_LN_6	LayerNorm	256 × 1536
28	Vision_Attn_6	Multi-Head Attention	256 × 1536
29	Vision_Add_6	Add	256 × 1536
30	Vision_FFN_6	Feed Forward	256 × 1536
31	Vision_LN_7	LayerNorm	256 × 1536
32	Vision_Attn_7	Multi-Head Attention	256 × 1536
33	Vision_Add_7	Add	256 × 1536
34	Vision_FFN_7	Feed Forward	256 × 1536
35	Vision_LN_8	LayerNorm	256 × 1536
36	Vision_Attn_8	Multi-Head Attention	256 × 1536
37	Vision_Add_8	Add	256 × 1536
38	Vision_FFN_8	Feed Forward	256 × 1536
39	Vision_LN_9	LayerNorm	256 × 1536
40	Vision_Attn_9	Multi-Head Attention	256 × 1536
41	Vision_Add_9	Add	256 × 1536
42	Vision_FFN_9	Feed Forward	256 × 1536
43	Vision_LN_10	LayerNorm	256 × 1536
44	Vision_Attn_10	Multi-Head Attention	256 × 1536
45	Vision_Add_10	Add	256 × 1536
46	Vision_FFN_10	Feed Forward	256 × 1536
47	Vision_LN_11	LayerNorm	256 × 1536
48	Vision_Attn_11	Multi-Head Attention	256 × 1536
49	Vision_Add_11	Add	256 × 1536
50	Vision_FFN_11	Feed Forward	256 × 1536
51	Vision_LN_12	LayerNorm	256 × 1536
52	Vision_Attn_12	Multi-Head Attention	256 × 1536
53	Vision_Add_12	Add	256 × 1536
54	Vision_FFN_12	Feed Forward	256 × 1536
55	Vision_LN_13	LayerNorm	256 × 1536
56	Vision_Attn_13	Multi-Head Attention	256 × 1536
57	Vision_Add_13	Add	256 × 1536
58	Vision_FFN_13	Feed Forward	256 × 1536
59	Vision_LN_14	LayerNorm	256 × 1536
60	Vision_Attn_14	Multi-Head Attention	256 × 1536
61	Vision_Add_14	Add	256 × 1536
62	Vision_FFN_14	Feed Forward	256 × 1536
63	Vision_LN_15	LayerNorm	256 × 1536
64	Vision_Attn_15	Multi-Head Attention	256 × 1536
65	Vision_Add_15	Add	256 × 1536
66	Vision_FFN_15	Feed Forward	256 × 1536
67	Vision_LN_16	LayerNorm	256 × 1536
68	Vision_Attn_16	Multi-Head Attention	256 × 1536
69	Vision_Add_16	Add	256 × 1536
70	Vision_FFN_16	Feed Forward	256 × 1536
71	Vision_LN_17	LayerNorm	256 × 1536
72	Vision_Attn_17	Multi-Head Attention	256 × 1536
73	Vision_Add_17	Add	256 × 1536
74	Vision_FFN_17	Feed Forward	256 × 1536
75	Vision_LN_18	LayerNorm	256 × 1536
76	Vision_Attn_18	Multi-Head Attention	256 × 1536
77	Vision_Add_18	Add	256 × 1536
78	Vision_FFN_18	Feed Forward	256 × 1536
79	Vision_LN_19	LayerNorm	256 × 1536
80	Vision_Attn_19	Multi-Head Attention	256 × 1536
81	Vision_Add_19	Add	256 × 1536
82	Vision_FFN_19	Feed Forward	256 × 1536
83	Vision_LN_20	LayerNorm	256 × 1536
84	Vision_Attn_20	Multi-Head Attention	256 × 1536
85	Vision_Add_20	Add	256 × 1536
86	Vision_FFN_20	Feed Forward	256 × 1536
87	Vision_LN_21	LayerNorm	256 × 1536
88	Vision_Attn_21	Multi-Head Attention	256 × 1536
89	Vision_Add_21	Add	256 × 1536
90	Vision_FFN_21	Feed Forward	256 × 1536
91	Vision_LN_22	LayerNorm	256 × 1536
92	Vision_Attn_22	Multi-Head Attention	256 × 1536
93	Vision_Add_22	Add	256 × 1536
94	Vision_FFN_22	Feed Forward	256 × 1536
95	Vision_LN_23	LayerNorm	256 × 1536
96	Vision_Attn_23	Multi-Head Attention	256 × 1536
97	Vision_Add_23	Add	256 × 1536
98	Vision_FFN_23	Feed Forward	256 × 1536
99	Vision_LN_24	LayerNorm	256 × 1536
100	Vision_Attn_24	Multi-Head Attention	256 × 1536
101	Vision_Add_24	Add	256 × 1536
102	Vision_FFN_24	Feed Forward	256 × 1536
103	Vision_LN_25	LayerNorm	256 × 1536
104	Vision_Attn_25	Multi-Head Attention	256 × 1536
105	Vision_Add_25	Add	256 × 1536
106	Vision_FFN_25	Feed Forward	256 × 1536
107	Vision_LN_26	LayerNorm	256 × 1536
108	Vision_Attn_26	Multi-Head Attention	256 × 1536
109	Vision_Add_26	Add	256 × 1536
110	Vision_FFN_26	Feed Forward	256 × 1536
111	Vision_LN_27	LayerNorm	256 × 1536
112	Vision_Attn_27	Multi-Head Attention	256 × 1536
113	Vision_Add_27	Add	256 × 1536
114	Vision_FFN_27	Feed Forward	256 × 1536
115	Vision_LN_28	LayerNorm	256 × 1536
116	Vision_Attn_28	Multi-Head Attention	256 × 1536
117	Vision_Add_28	Add	256 × 1536
118	Vision_FFN_28	Feed Forward	256 × 1536
119	Vision_LN_29	LayerNorm	256 × 1536
120	Vision_Attn_29	Multi-Head Attention	256 × 1536
121	Vision_Add_29	Add	256 × 1536
122	Vision_FFN_29	Feed Forward	256 × 1536
123	Vision_LN_30	LayerNorm	256 × 1536
124	Vision_Attn_30	Multi-Head Attention	256 × 1536
125	Vision_Add_30	Add	256 × 1536
126	Vision_FFN_30	Feed Forward	256 × 1536
127	Vision_LN_31	LayerNorm	256 × 1536
128	Vision_Attn_31	Multi-Head Attention	256 × 1536
129	Vision_Add_31	Add	256 × 1536
130	Vision_FFN_31	Feed Forward	256 × 1536
131	Vision_LN_32	LayerNorm	256 × 1536
132	Vision_Attn_32	Multi-Head Attention	256 × 1536
133	Vision_Add_32	Add	256 × 1536
134	Vision_FFN_32	Feed Forward	256 × 1536
135	Vision_LN_33	LayerNorm	256 × 1536
136	Vision_Attn_33	Multi-Head Attention	256 × 1536
137	Vision_Add_33	Add	256 × 1536
138	Vision_FFN_33	Feed Forward	256 × 1536
139	Vision_LN_34	LayerNorm	256 × 1536
140	Vision_Attn_34	Multi-Head Attention	256 × 1536
141	Vision_Add_34	Add	256 × 1536
142	Vision_FFN_34	Feed Forward	256 × 1536
143	Vision_LN_35	LayerNorm	256 × 1536
144	Vision_Attn_35	Multi-Head Attention	256 × 1536
145	Vision_Add_35	Add	256 × 1536
146	Vision_FFN_35	Feed Forward	256 × 1536
147	Vision_LN_36	LayerNorm	256 × 1536
148	Vision_Attn_36	Multi-Head Attention	256 × 1536
149	Vision_Add_36	Add	256 × 1536
150	Vision_FFN_36	Feed Forward	256 × 1536
151	Vision_LN_37	LayerNorm	256 × 1536
152	Vision_Attn_37	Multi-Head Attention	256 × 1536
153	Vision_Add_37	Add	256 × 1536
154	Vision_FFN_37	Feed Forward	256 × 1536
155	Vision_LN_38	LayerNorm	256 × 1536
156	Vision_Attn_38	Multi-Head Attention	256 × 1536
157	Vision_Add_38	Add	256 × 1536
158	Vision_FFN_38	Feed Forward	256 × 1536
159	Vision_LN_39	LayerNorm	256 × 1536
160	Vision_Attn_39	Multi-Head Attention	256 × 1536
161	Vision_Add_39	Add	256 × 1536
162	Vision_FFN_39	Feed Forward	256 × 1536
163	Vision_LN_40	LayerNorm	256 × 1536
164	Vision_Attn_40	Multi-Head Attention	256 × 1536
165	Vision_Add_40	Add	256 × 1536
166	Vision_FFN_40	Feed Forward	256 × 1536
167	Vision_LN_41	LayerNorm	256 × 1536
168	Vision_Attn_41	Multi-Head Attention	256 × 1536
169	Vision_Add_41	Add	256 × 1536
170	Vision_FFN_41	Feed Forward	256 × 1536
171	Vision_LN_42	LayerNorm	256 × 1536
172	Vision_Attn_42	Multi-Head Attention	256 × 1536
173	Vision_Add_42	Add	256 × 1536
174	Vision_FFN_42	Feed Forward	256 × 1536
175	Vision projector	Projection	256 × 1536
176	Vision tokens	Reshape	1 × 256 × 1536
177	Multimodal fusion (concat tokens)	Concatenate	1 × 131328 × 1536
178	RMSNorm_1_1	RMSNorm	1 × 131328 × 1536
179	Attention_1	Grouped Query Attn	1 × 131328 × 1536
180	Add_1_attn	Add	1 × 131328 × 1536
181	RMSNorm_1_2	RMSNorm	1 × 131328 × 1536
182	FFN_1	SwiGLU	1 × 131328 × 1536
183	Add_1_ffn	Add	1 × 131328 × 1536
184	RMSNorm_2_1	RMSNorm	1 × 131328 × 1536
185	Attention_2	Grouped Query Attn	1 × 131328 × 1536
186	Add_2_attn	Add	1 × 131328 × 1536
187	RMSNorm_2_2	RMSNorm	1 × 131328 × 1536
188	FFN_2	SwiGLU	1 × 131328 × 1536
189	Add_2_ffn	Add	1 × 131328 × 1536
190	RMSNorm_3_1	RMSNorm	1 × 131328 × 1536
191	Attention_3	Grouped Query Attn	1 × 131328 × 1536
192	Add_3_attn	Add	1 × 131328 × 1536
193	RMSNorm_3_2	RMSNorm	1 × 131328 × 1536
194	FFN_3	SwiGLU	1 × 131328 × 1536
195	Add_3_ffn	Add	1 × 131328 × 1536
196	RMSNorm_4_1	RMSNorm	1 × 131328 × 1536
197	Attention_4	Grouped Query Attn	1 × 131328 × 1536
198	Add_4_attn	Add	1 × 131328 × 1536
199	RMSNorm_4_2	RMSNorm	1 × 131328 × 1536
200	FFN_4	SwiGLU	1 × 131328 × 1536
201	Add_4_ffn	Add	1 × 131328 × 1536
202	RMSNorm_5_1	RMSNorm	1 × 131328 × 1536
203	Attention_5	Grouped Query Attn	1 × 131328 × 1536
204	Add_5_attn	Add	1 × 131328 × 1536
205	RMSNorm_5_2	RMSNorm	1 × 131328 × 1536
206	FFN_5	SwiGLU	1 × 131328 × 1536
207	Add_5_ffn	Add	1 × 131328 × 1536
208	RMSNorm_6_1	RMSNorm	1 × 131328 × 1536
209	Attention_6	Grouped Query Attn	1 × 131328 × 1536
210	Add_6_attn	Add	1 × 131328 × 1536
211	RMSNorm_6_2	RMSNorm	1 × 131328 × 1536
212	FFN_6	SwiGLU	1 × 131328 × 1536
213	Add_6_ffn	Add	1 × 131328 × 1536
214	RMSNorm_7_1	RMSNorm	1 × 131328 × 1536
215	Attention_7	Grouped Query Attn	1 × 131328 × 1536
216	Add_7_attn	Add	1 × 131328 × 1536
217	RMSNorm_7_2	RMSNorm	1 × 131328 × 1536
218	FFN_7	SwiGLU	1 × 131328 × 1536
219	Add_7_ffn	Add	1 × 131328 × 1536
220	RMSNorm_8_1	RMSNorm	1 × 131328 × 1536
221	Attention_8	Grouped Query Attn	1 × 131328 × 1536
222	Add_8_attn	Add	1 × 131328 × 1536
223	RMSNorm_8_2	RMSNorm	1 × 131328 × 1536
224	FFN_8	SwiGLU	1 × 131328 × 1536
225	Add_8_ffn	Add	1 × 131328 × 1536
226	RMSNorm_9_1	RMSNorm	1 × 131328 × 1536
227	Attention_9	Grouped Query Attn	1 × 131328 × 1536
228	Add_9_attn	Add	1 × 131328 × 1536
229	RMSNorm_9_2	RMSNorm	1 × 131328 × 1536
230	FFN_9	SwiGLU	1 × 131328 × 1536
231	Add_9_ffn	Add	1 × 131328 × 1536
232	RMSNorm_10_1	RMSNorm	1 × 131328 × 1536
233	Attention_10	Grouped Query Attn	1 × 131328 × 1536
234	Add_10_attn	Add	1 × 131328 × 1536
235	RMSNorm_10_2	RMSNorm	1 × 131328 × 1536
236	FFN_10	SwiGLU	1 × 131328 × 1536
237	Add_10_ffn	Add	1 × 131328 × 1536
238	RMSNorm_11_1	RMSNorm	1 × 131328 × 1536
239	Attention_11	Grouped Query Attn	1 × 131328 × 1536
240	Add_11_attn	Add	1 × 131328 × 1536
241	RMSNorm_11_2	RMSNorm	1 × 131328 × 1536
242	FFN_11	SwiGLU	1 × 131328 × 1536
243	Add_11_ffn	Add	1 × 131328 × 1536
244	RMSNorm_12_1	RMSNorm	1 × 131328 × 1536
245	Attention_12	Grouped Query Attn	1 × 131328 × 1536
246	Add_12_attn	Add	1 × 131328 × 1536
247	RMSNorm_12_2	RMSNorm	1 × 131328 × 1536
248	FFN_12	SwiGLU	1 × 131328 × 1536
249	Add_12_ffn	Add	1 × 131328 × 1536
250	RMSNorm_13_1	RMSNorm	1 × 131328 × 1536
251	Attention_13	Grouped Query Attn	1 × 131328 × 1536
252	Add_13_attn	Add	1 × 131328 × 1536
253	RMSNorm_13_2	RMSNorm	1 × 131328 × 1536
254	FFN_13	SwiGLU	1 × 131328 × 1536
255	Add_13_ffn	Add	1 × 131328 × 1536
256	RMSNorm_14_1	RMSNorm	1 × 131328 × 1536
257	Attention_14	Grouped Query Attn	1 × 131328 × 1536
258	Add_14_attn	Add	1 × 131328 × 1536
259	RMSNorm_14_2	RMSNorm	1 × 131328 × 1536
260	FFN_14	SwiGLU	1 × 131328 × 1536
261	Add_14_ffn	Add	1 × 131328 × 1536
262	RMSNorm_15_1	RMSNorm	1 × 131328 × 1536
263	Attention_15	Grouped Query Attn	1 × 131328 × 1536
264	Add_15_attn	Add	1 × 131328 × 1536
265	RMSNorm_15_2	RMSNorm	1 × 131328 × 1536
266	FFN_15	SwiGLU	1 × 131328 × 1536
267	Add_15_ffn	Add	1 × 131328 × 1536
268	RMSNorm_16_1	RMSNorm	1 × 131328 × 1536
269	Attention_16	Grouped Query Attn	1 × 131328 × 1536
270	Add_16_attn	Add	1 × 131328 × 1536
271	RMSNorm_16_2	RMSNorm	1 × 131328 × 1536
272	FFN_16	SwiGLU	1 × 131328 × 1536
273	Add_16_ffn	Add	1 × 131328 × 1536
274	RMSNorm_17_1	RMSNorm	1 × 131328 × 1536
275	Attention_17	Grouped Query Attn	1 × 131328 × 1536
276	Add_17_attn	Add	1 × 131328 × 1536
277	RMSNorm_17_2	RMSNorm	1 × 131328 × 1536
278	FFN_17	SwiGLU	1 × 131328 × 1536
279	Add_17_ffn	Add	1 × 131328 × 1536
280	RMSNorm_18_1	RMSNorm	1 × 131328 × 1536
281	Attention_18	Grouped Query Attn	1 × 131328 × 1536
282	Add_18_attn	Add	1 × 131328 × 1536
283	RMSNorm_18_2	RMSNorm	1 × 131328 × 1536
284	FFN_18	SwiGLU	1 × 131328 × 1536
285	Add_18_ffn	Add	1 × 131328 × 1536
286	RMSNorm_19_1	RMSNorm	1 × 131328 × 1536
287	Attention_19	Grouped Query Attn	1 × 131328 × 1536
288	Add_19_attn	Add	1 × 131328 × 1536
289	RMSNorm_19_2	RMSNorm	1 × 131328 × 1536
290	FFN_19	SwiGLU	1 × 131328 × 1536
291	Add_19_ffn	Add	1 × 131328 × 1536
292	RMSNorm_20_1	RMSNorm	1 × 131328 × 1536
293	Attention_20	Grouped Query Attn	1 × 131328 × 1536
294	Add_20_attn	Add	1 × 131328 × 1536
295	RMSNorm_20_2	RMSNorm	1 × 131328 × 1536
296	FFN_20	SwiGLU	1 × 131328 × 1536
297	Add_20_ffn	Add	1 × 131328 × 1536
298	RMSNorm_21_1	RMSNorm	1 × 131328 × 1536
299	Attention_21	Grouped Query Attn	1 × 131328 × 1536
300	Add_21_attn	Add	1 × 131328 × 1536
301	RMSNorm_21_2	RMSNorm	1 × 131328 × 1536
302	FFN_21	SwiGLU	1 × 131328 × 1536
303	Add_21_ffn	Add	1 × 131328 × 1536
304	RMSNorm_22_1	RMSNorm	1 × 131328 × 1536
305	Attention_22	Grouped Query Attn	1 × 131328 × 1536
306	Add_22_attn	Add	1 × 131328 × 1536
307	RMSNorm_22_2	RMSNorm	1 × 131328 × 1536
308	FFN_22	SwiGLU	1 × 131328 × 1536
309	Add_22_ffn	Add	1 × 131328 × 1536
310	RMSNorm_23_1	RMSNorm	1 × 131328 × 1536
311	Attention_23	Grouped Query Attn	1 × 131328 × 1536
312	Add_23_attn	Add	1 × 131328 × 1536
313	RMSNorm_23_2	RMSNorm	1 × 131328 × 1536
314	FFN_23	SwiGLU	1 × 131328 × 1536
315	Add_23_ffn	Add	1 × 131328 × 1536
316	RMSNorm_24_1	RMSNorm	1 × 131328 × 1536
317	Attention_24	Grouped Query Attn	1 × 131328 × 1536
318	Add_24_attn	Add	1 × 131328 × 1536
319	RMSNorm_24_2	RMSNorm	1 × 131328 × 1536
320	FFN_24	SwiGLU	1 × 131328 × 1536
321	Add_24_ffn	Add	1 × 131328 × 1536
322	RMSNorm_25_1	RMSNorm	1 × 131328 × 1536
323	Attention_25	Grouped Query Attn	1 × 131328 × 1536
324	Add_25_attn	Add	1 × 131328 × 1536
325	RMSNorm_25_2	RMSNorm	1 × 131328 × 1536
326	FFN_25	SwiGLU	1 × 131328 × 1536
327	Add_25_ffn	Add	1 × 131328 × 1536
328	RMSNorm_26_1	RMSNorm	1 × 131328 × 1536
329	Attention_26	Grouped Query Attn	1 × 131328 × 1536
330	Add_26_attn	Add	1 × 131328 × 1536
331	RMSNorm_26_2	RMSNorm	1 × 131328 × 1536
332	FFN_26	SwiGLU	1 × 131328 × 1536
333	Add_26_ffn	Add	1 × 131328 × 1536
334	RMSNorm_27_1	RMSNorm	1 × 131328 × 1536
335	Attention_27	Grouped Query Attn	1 × 131328 × 1536
336	Add_27_attn	Add	1 × 131328 × 1536
337	RMSNorm_27_2	RMSNorm	1 × 131328 × 1536
338	FFN_27	SwiGLU	1 × 131328 × 1536
339	Add_27_ffn	Add	1 × 131328 × 1536
340	RMSNorm_28_1	RMSNorm	1 × 131328 × 1536
341	Attention_28	Grouped Query Attn	1 × 131328 × 1536
342	Add_28_attn	Add	1 × 131328 × 1536
343	RMSNorm_28_2	RMSNorm	1 × 131328 × 1536
344	FFN_28	SwiGLU	1 × 131328 × 1536
345	Add_28_ffn	Add	1 × 131328 × 1536
346	Final_RMSNorm	RMSNorm	1 × 131328 × 1536
347	LM_Head	Linear	1 × 131328 × 151936
348	Output	Output	1 × 131328 × 151936

What the verifier says

warn"RoPE" receives input but its output is not connected. This layer will be unreachable in the forward pass. Fix: Connect the output forward, or add an Output node if this is the final layer.
dead-end

infoLLaMA uses intermediateSize ≈ ⌊(8/3 × D) / 256⌋ × 256. Current: 8960 (5.83× embedDim). Expected: ~4096. Fix: Set intermediateSize to 4096 for embedDim=1536.
swiglu-dim-convention

infoAt 70 stacked attention layers, residual-branch outputs add up; unscaled init lets activation variance grow with depth. GPT-2/LLaMA-family models scale the residual projections by depth (N(0, 0.02 / √(2L))). Fix: Scale residual output projections by depth: nn.init.normal_(w, std=0.02 / math.sqrt(2 * n_layers))
deep-attention-default-init

Do this to your own model

Same numbers, on a model in your repo, in one command. No account.

pip install neurarch-trace
neurarch-trace dots-studio/dots.mocr --plan --share

Paste a model instead Markdown twin Index JSON POST /api/v1/plan

Other dots_ocr checkpoints

dots.ocr

2.72B derived · -10.4% against the checkpoint