descriptions
Browse files- zephyr_f32_int2.txt +650 -0
- zephyr_f32_int3.txt +326 -0
- zephyr_int8_int3.txt +362 -0
zephyr_f32_int2.txt
ADDED
|
@@ -0,0 +1,650 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
(dream) tb@IBM-PF38WZKF:~/funstreams/AI$ ./llama.cpp/quantize zephyr_f32.gguf Q3_K
|
| 2 |
+
main: build = 1798 (128de35)
|
| 3 |
+
main: built with cc (Ubuntu 9.4.0-1ubuntu1~20.04.2) 9.4.0 for x86_64-linux-gnu
|
| 4 |
+
main: quantizing 'zephyr_f32.gguf' to 'ggml-model-Q3_K.gguf' as Q3_K
|
| 5 |
+
llama_model_loader: loaded meta data with 22 key-value pairs and 291 tensors from zephyr_f32.gguf (version GGUF V3 (latest))
|
| 6 |
+
llama_model_loader: Dumping metadata keys/values. Note: KV overrides do not apply in this output.
|
| 7 |
+
llama_model_loader: - kv 0: general.architecture str = llama
|
| 8 |
+
llama_model_loader: - kv 1: general.name str = .
|
| 9 |
+
llama_model_loader: - kv 2: llama.context_length u32 = 32768
|
| 10 |
+
llama_model_loader: - kv 3: llama.embedding_length u32 = 4096
|
| 11 |
+
llama_model_loader: - kv 4: llama.block_count u32 = 32
|
| 12 |
+
llama_model_loader: - kv 5: llama.feed_forward_length u32 = 14336
|
| 13 |
+
llama_model_loader: - kv 6: llama.rope.dimension_count u32 = 128
|
| 14 |
+
llama_model_loader: - kv 7: llama.attention.head_count u32 = 32
|
| 15 |
+
llama_model_loader: - kv 8: llama.attention.head_count_kv u32 = 8
|
| 16 |
+
llama_model_loader: - kv 9: llama.attention.layer_norm_rms_epsilon f32 = 0.000010
|
| 17 |
+
llama_model_loader: - kv 10: llama.rope.freq_base f32 = 10000.000000
|
| 18 |
+
llama_model_loader: - kv 11: general.file_type u32 = 0
|
| 19 |
+
llama_model_loader: - kv 12: tokenizer.ggml.model str = llama
|
| 20 |
+
llama_model_loader: - kv 13: tokenizer.ggml.tokens arr[str,32000] = ["<unk>", "<s>", "</s>", "<0x00>", "<...
|
| 21 |
+
llama_model_loader: - kv 14: tokenizer.ggml.scores arr[f32,32000] = [0.000000, 0.000000, 0.000000, 0.0000...
|
| 22 |
+
llama_model_loader: - kv 15: tokenizer.ggml.token_type arr[i32,32000] = [2, 3, 3, 6, 6, 6, 6, 6, 6, 6, 6, 6, ...
|
| 23 |
+
llama_model_loader: - kv 16: tokenizer.ggml.merges arr[str,58980] = ["▁ t", "i n", "e r", "▁ a", "h e...
|
| 24 |
+
llama_model_loader: - kv 17: tokenizer.ggml.bos_token_id u32 = 1
|
| 25 |
+
llama_model_loader: - kv 18: tokenizer.ggml.eos_token_id u32 = 2
|
| 26 |
+
llama_model_loader: - kv 19: tokenizer.ggml.unknown_token_id u32 = 0
|
| 27 |
+
llama_model_loader: - kv 20: tokenizer.ggml.padding_token_id u32 = 2
|
| 28 |
+
llama_model_loader: - kv 21: tokenizer.chat_template str = {% for message in messages %}\n{% if m...
|
| 29 |
+
llama_model_loader: - type f32: 291 tensors
|
| 30 |
+
llama_model_quantize_internal: meta size = 1671648 bytes
|
| 31 |
+
[ 1/ 291] token_embd.weight - [ 4096, 32000, 1, 1], type = f32, quantizing to q3_K .. size = 500.00 MiB -> 53.71 MiB | hist:
|
| 32 |
+
[ 2/ 291] blk.0.attn_norm.weight - [ 4096, 1, 1, 1], type = f32, size = 0.016 MB
|
| 33 |
+
[ 3/ 291] blk.0.ffn_down.weight - [14336, 4096, 1, 1], type = f32, quantizing to q5_K .. size = 224.00 MiB -> 38.50 MiB | hist:
|
| 34 |
+
[ 4/ 291] blk.0.ffn_gate.weight - [ 4096, 14336, 1, 1], type = f32, quantizing to q3_K .. size = 224.00 MiB -> 24.06 MiB | hist:
|
| 35 |
+
[ 5/ 291] blk.0.ffn_up.weight - [ 4096, 14336, 1, 1], type = f32, quantizing to q3_K .. size = 224.00 MiB -> 24.06 MiB | hist:
|
| 36 |
+
[ 6/ 291] blk.0.ffn_norm.weight - [ 4096, 1, 1, 1], type = f32, size = 0.016 MB
|
| 37 |
+
[ 7/ 291] blk.0.attn_k.weight - [ 4096, 1024, 1, 1], type = f32, quantizing to q3_K .. size = 16.00 MiB -> 1.72 MiB | hist:
|
| 38 |
+
[ 8/ 291] blk.0.attn_output.weight - [ 4096, 4096, 1, 1], type = f32, quantizing to q4_K .. size = 64.00 MiB -> 9.00 MiB | hist:
|
| 39 |
+
[ 9/ 291] blk.0.attn_q.weight - [ 4096, 4096, 1, 1], type = f32, quantizing to q3_K .. size = 64.00 MiB -> 6.88 MiB | hist:
|
| 40 |
+
[ 10/ 291] blk.0.attn_v.weight - [ 4096, 1024, 1, 1], type = f32, quantizing to q5_K .. size = 16.00 MiB -> 2.75 MiB | hist:
|
| 41 |
+
[ 11/ 291] blk.1.attn_norm.weight - [ 4096, 1, 1, 1], type = f32, size = 0.016 MB
|
| 42 |
+
[ 12/ 291] blk.1.ffn_down.weight - [14336, 4096, 1, 1], type = f32, quantizing to q5_K .. size = 224.00 MiB -> 38.50 MiB | hist:
|
| 43 |
+
[ 13/ 291] blk.1.ffn_gate.weight - [ 4096, 14336, 1, 1], type = f32, quantizing to q3_K .. size = 224.00 MiB -> 24.06 MiB | hist:
|
| 44 |
+
[ 14/ 291] blk.1.ffn_up.weight - [ 4096, 14336, 1, 1], type = f32, quantizing to q3_K .. size = 224.00 MiB -> 24.06 MiB | hist:
|
| 45 |
+
[ 15/ 291] blk.1.ffn_norm.weight - [ 4096, 1, 1, 1], type = f32, size = 0.016 MB
|
| 46 |
+
[ 16/ 291] blk.1.attn_k.weight - [ 4096, 1024, 1, 1], type = f32, quantizing to q3_K .. size = 16.00 MiB -> 1.72 MiB | hist:
|
| 47 |
+
[ 17/ 291] blk.1.attn_output.weight - [ 4096, 4096, 1, 1], type = f32, quantizing to q4_K .. size = 64.00 MiB -> 9.00 MiB | hist:
|
| 48 |
+
[ 18/ 291] blk.1.attn_q.weight - [ 4096, 4096, 1, 1], type = f32, quantizing to q3_K .. size = 64.00 MiB -> 6.88 MiB | hist:
|
| 49 |
+
[ 19/ 291] blk.1.attn_v.weight - [ 4096, 1024, 1, 1], type = f32, quantizing to q5_K .. size = 16.00 MiB -> 2.75 MiB | hist:
|
| 50 |
+
[ 20/ 291] blk.2.attn_norm.weight - [ 4096, 1, 1, 1], type = f32, size = 0.016 MB
|
| 51 |
+
[ 21/ 291] blk.2.ffn_down.weight - [14336, 4096, 1, 1], type = f32, quantizing to q4_K .. size = 224.00 MiB -> 31.50 MiB | hist:
|
| 52 |
+
[ 22/ 291] blk.2.ffn_gate.weight - [ 4096, 14336, 1, 1], type = f32, quantizing to q3_K .. size = 224.00 MiB -> 24.06 MiB | hist:
|
| 53 |
+
[ 23/ 291] blk.2.ffn_up.weight - [ 4096, 14336, 1, 1], type = f32, quantizing to q3_K .. size = 224.00 MiB -> 24.06 MiB | hist:
|
| 54 |
+
[ 24/ 291] blk.2.ffn_norm.weight - [ 4096, 1, 1, 1], type = f32, size = 0.016 MB
|
| 55 |
+
[ 25/ 291] blk.2.attn_k.weight - [ 4096, 1024, 1, 1], type = f32, quantizing to q3_K .. size = 16.00 MiB -> 1.72 MiB | hist:
|
| 56 |
+
[ 26/ 291] blk.2.attn_output.weight - [ 4096, 4096, 1, 1], type = f32, quantizing to q4_K .. size = 64.00 MiB -> 9.00 MiB | hist:
|
| 57 |
+
[ 27/ 291] blk.2.attn_q.weight - [ 4096, 4096, 1, 1], type = f32, quantizing to q3_K .. size = 64.00 MiB -> 6.88 MiB | hist:
|
| 58 |
+
[ 28/ 291] blk.2.attn_v.weight - [ 4096, 1024, 1, 1], type = f32, quantizing to q4_K .. size = 16.00 MiB -> 2.25 MiB | hist:
|
| 59 |
+
[ 29/ 291] blk.3.ffn_gate.weight - [ 4096, 14336, 1, 1], type = f32, quantizing to q3_K .. size = 224.00 MiB -> 24.06 MiB | hist:
|
| 60 |
+
[ 30/ 291] blk.3.ffn_up.weight - [ 4096, 14336, 1, 1], type = f32, quantizing to q3_K .. size = 224.00 MiB -> 24.06 MiB | hist:
|
| 61 |
+
[ 31/ 291] blk.3.attn_k.weight - [ 4096, 1024, 1, 1], type = f32, quantizing to q3_K .. size = 16.00 MiB -> 1.72 MiB | hist:
|
| 62 |
+
[ 32/ 291] blk.3.attn_output.weight - [ 4096, 4096, 1, 1], type = f32, quantizing to q4_K .. size = 64.00 MiB -> 9.00 MiB | hist:
|
| 63 |
+
[ 33/ 291] blk.3.attn_q.weight - [ 4096, 4096, 1, 1], type = f32, quantizing to q3_K .. size = 64.00 MiB -> 6.88 MiB | hist:
|
| 64 |
+
[ 34/ 291] blk.3.attn_v.weight - [ 4096, 1024, 1, 1], type = f32, quantizing to q4_K .. size = 16.00 MiB -> 2.25 MiB | hist:
|
| 65 |
+
[ 35/ 291] blk.3.attn_norm.weight - [ 4096, 1, 1, 1], type = f32, size = 0.016 MB
|
| 66 |
+
[ 36/ 291] blk.3.ffn_down.weight - [14336, 4096, 1, 1], type = f32, quantizing to q4_K .. size = 224.00 MiB -> 31.50 MiB | hist:
|
| 67 |
+
[ 37/ 291] blk.3.ffn_norm.weight - [ 4096, 1, 1, 1], type = f32, size = 0.016 MB
|
| 68 |
+
[ 38/ 291] blk.4.attn_norm.weight - [ 4096, 1, 1, 1], type = f32, size = 0.016 MB
|
| 69 |
+
[ 39/ 291] blk.4.ffn_down.weight - [14336, 4096, 1, 1], type = f32, quantizing to q4_K .. size = 224.00 MiB -> 31.50 MiB | hist:
|
| 70 |
+
[ 40/ 291] blk.4.ffn_gate.weight - [ 4096, 14336, 1, 1], type = f32, quantizing to q3_K .. size = 224.00 MiB -> 24.06 MiB | hist:
|
| 71 |
+
[ 41/ 291] blk.4.ffn_up.weight - [ 4096, 14336, 1, 1], type = f32, quantizing to q3_K .. size = 224.00 MiB -> 24.06 MiB | hist:
|
| 72 |
+
[ 42/ 291] blk.4.ffn_norm.weight - [ 4096, 1, 1, 1], type = f32, size = 0.016 MB
|
| 73 |
+
[ 43/ 291] blk.4.attn_k.weight - [ 4096, 1024, 1, 1], type = f32, quantizing to q3_K .. size = 16.00 MiB -> 1.72 MiB | hist:
|
| 74 |
+
[ 44/ 291] blk.4.attn_output.weight - [ 4096, 4096, 1, 1], type = f32, quantizing to q4_K .. size = 64.00 MiB -> 9.00 MiB | hist:
|
| 75 |
+
[ 45/ 291] blk.4.attn_q.weight - [ 4096, 4096, 1, 1], type = f32, quantizing to q3_K .. size = 64.00 MiB -> 6.88 MiB | hist:
|
| 76 |
+
[ 46/ 291] blk.4.attn_v.weight - [ 4096, 1024, 1, 1], type = f32, quantizing to q4_K .. size = 16.00 MiB -> 2.25 MiB | hist:
|
| 77 |
+
[ 47/ 291] blk.5.attn_norm.weight - [ 4096, 1, 1, 1], type = f32, size = 0.016 MB
|
| 78 |
+
[ 48/ 291] blk.5.ffn_down.weight - [14336, 4096, 1, 1], type = f32, quantizing to q4_K .. size = 224.00 MiB -> 31.50 MiB | hist:
|
| 79 |
+
[ 49/ 291] blk.5.ffn_gate.weight - [ 4096, 14336, 1, 1], type = f32, quantizing to q3_K .. size = 224.00 MiB -> 24.06 MiB | hist:
|
| 80 |
+
[ 50/ 291] blk.5.ffn_up.weight - [ 4096, 14336, 1, 1], type = f32, quantizing to q3_K .. size = 224.00 MiB -> 24.06 MiB | hist:
|
| 81 |
+
[ 51/ 291] blk.5.ffn_norm.weight - [ 4096, 1, 1, 1], type = f32, size = 0.016 MB
|
| 82 |
+
[ 52/ 291] blk.5.attn_k.weight - [ 4096, 1024, 1, 1], type = f32, quantizing to q3_K .. size = 16.00 MiB -> 1.72 MiB | hist:
|
| 83 |
+
[ 53/ 291] blk.5.attn_output.weight - [ 4096, 4096, 1, 1], type = f32, quantizing to q4_K .. size = 64.00 MiB -> 9.00 MiB | hist:
|
| 84 |
+
[ 54/ 291] blk.5.attn_q.weight - [ 4096, 4096, 1, 1], type = f32, quantizing to q3_K .. size = 64.00 MiB -> 6.88 MiB | hist:
|
| 85 |
+
[ 55/ 291] blk.5.attn_v.weight - [ 4096, 1024, 1, 1], type = f32, quantizing to q4_K .. size = 16.00 MiB -> 2.25 MiB | hist:
|
| 86 |
+
[ 56/ 291] blk.6.attn_norm.weight - [ 4096, 1, 1, 1], type = f32, size = 0.016 MB
|
| 87 |
+
[ 57/ 291] blk.6.ffn_down.weight - [14336, 4096, 1, 1], type = f32, quantizing to q4_K .. size = 224.00 MiB -> 31.50 MiB | hist:
|
| 88 |
+
[ 58/ 291] blk.6.ffn_gate.weight - [ 4096, 14336, 1, 1], type = f32, quantizing to q3_K .. size = 224.00 MiB -> 24.06 MiB | hist:
|
| 89 |
+
[ 59/ 291] blk.6.ffn_up.weight - [ 4096, 14336, 1, 1], type = f32, quantizing to q3_K .. size = 224.00 MiB -> 24.06 MiB | hist:
|
| 90 |
+
[ 60/ 291] blk.6.ffn_norm.weight - [ 4096, 1, 1, 1], type = f32, size = 0.016 MB
|
| 91 |
+
[ 61/ 291] blk.6.attn_k.weight - [ 4096, 1024, 1, 1], type = f32, quantizing to q3_K .. size = 16.00 MiB -> 1.72 MiB | hist:
|
| 92 |
+
[ 62/ 291] blk.6.attn_output.weight - [ 4096, 4096, 1, 1], type = f32, quantizing to q4_K .. size = 64.00 MiB -> 9.00 MiB | hist:
|
| 93 |
+
[ 63/ 291] blk.6.attn_q.weight - [ 4096, 4096, 1, 1], type = f32, quantizing to q3_K .. size = 64.00 MiB -> 6.88 MiB | hist:
|
| 94 |
+
[ 64/ 291] blk.6.attn_v.weight - [ 4096, 1024, 1, 1], type = f32, quantizing to q4_K .. size = 16.00 MiB -> 2.25 MiB | hist:
|
| 95 |
+
[ 65/ 291] blk.7.attn_norm.weight - [ 4096, 1, 1, 1], type = f32, size = 0.016 MB
|
| 96 |
+
[ 66/ 291] blk.7.ffn_down.weight - [14336, 4096, 1, 1], type = f32, quantizing to q4_K .. size = 224.00 MiB -> 31.50 MiB | hist:
|
| 97 |
+
[ 67/ 291] blk.7.ffn_gate.weight - [ 4096, 14336, 1, 1], type = f32, quantizing to q3_K .. size = 224.00 MiB -> 24.06 MiB | hist:
|
| 98 |
+
[ 68/ 291] blk.7.ffn_up.weight - [ 4096, 14336, 1, 1], type = f32, quantizing to q3_K .. size = 224.00 MiB -> 24.06 MiB | hist:
|
| 99 |
+
[ 69/ 291] blk.7.ffn_norm.weight - [ 4096, 1, 1, 1], type = f32, size = 0.016 MB
|
| 100 |
+
[ 70/ 291] blk.7.attn_k.weight - [ 4096, 1024, 1, 1], type = f32, quantizing to q3_K .. size = 16.00 MiB -> 1.72 MiB | hist:
|
| 101 |
+
[ 71/ 291] blk.7.attn_output.weight - [ 4096, 4096, 1, 1], type = f32, quantizing to q4_K .. size = 64.00 MiB -> 9.00 MiB | hist:
|
| 102 |
+
[ 72/ 291] blk.7.attn_q.weight - [ 4096, 4096, 1, 1], type = f32, quantizing to q3_K .. size = 64.00 MiB -> 6.88 MiB | hist:
|
| 103 |
+
[ 73/ 291] blk.7.attn_v.weight - [ 4096, 1024, 1, 1], type = f32, quantizing to q4_K .. size = 16.00 MiB -> 2.25 MiB | hist:
|
| 104 |
+
[ 74/ 291] blk.8.attn_k.weight - [ 4096, 1024, 1, 1], type = f32, quantizing to q3_K .. size = 16.00 MiB -> 1.72 MiB | hist:
|
| 105 |
+
[ 75/ 291] blk.8.attn_output.weight - [ 4096, 4096, 1, 1], type = f32, quantizing to q4_K .. size = 64.00 MiB -> 9.00 MiB | hist:
|
| 106 |
+
[ 76/ 291] blk.8.attn_q.weight - [ 4096, 4096, 1, 1], type = f32, quantizing to q3_K .. size = 64.00 MiB -> 6.88 MiB | hist:
|
| 107 |
+
[ 77/ 291] blk.8.attn_v.weight - [ 4096, 1024, 1, 1], type = f32, quantizing to q4_K .. size = 16.00 MiB -> 2.25 MiB | hist:
|
| 108 |
+
[ 78/ 291] blk.10.attn_norm.weight - [ 4096, 1, 1, 1], type = f32, size = 0.016 MB
|
| 109 |
+
[ 79/ 291] blk.10.ffn_down.weight - [14336, 4096, 1, 1], type = f32, quantizing to q4_K .. size = 224.00 MiB -> 31.50 MiB | hist:
|
| 110 |
+
[ 80/ 291] blk.10.ffn_gate.weight - [ 4096, 14336, 1, 1], type = f32, quantizing to q3_K .. size = 224.00 MiB -> 24.06 MiB | hist:
|
| 111 |
+
[ 81/ 291] blk.10.ffn_up.weight - [ 4096, 14336, 1, 1], type = f32, quantizing to q3_K .. size = 224.00 MiB -> 24.06 MiB | hist:
|
| 112 |
+
[ 82/ 291] blk.10.ffn_norm.weight - [ 4096, 1, 1, 1], type = f32, size = 0.016 MB
|
| 113 |
+
[ 83/ 291] blk.10.attn_k.weight - [ 4096, 1024, 1, 1], type = f32, quantizing to q3_K .. size = 16.00 MiB -> 1.72 MiB | hist:
|
| 114 |
+
[ 84/ 291] blk.10.attn_output.weight - [ 4096, 4096, 1, 1], type = f32, quantizing to q4_K .. size = 64.00 MiB -> 9.00 MiB | hist:
|
| 115 |
+
[ 85/ 291] blk.10.attn_q.weight - [ 4096, 4096, 1, 1], type = f32, quantizing to q3_K .. size = 64.00 MiB -> 6.88 MiB | hist:
|
| 116 |
+
[ 86/ 291] blk.10.attn_v.weight - [ 4096, 1024, 1, 1], type = f32, quantizing to q4_K .. size = 16.00 MiB -> 2.25 MiB | hist:
|
| 117 |
+
[ 87/ 291] blk.11.attn_norm.weight - [ 4096, 1, 1, 1], type = f32, size = 0.016 MB
|
| 118 |
+
[ 88/ 291] blk.11.ffn_down.weight - [14336, 4096, 1, 1], type = f32, quantizing to q4_K .. size = 224.00 MiB -> 31.50 MiB | hist:
|
| 119 |
+
[ 89/ 291] blk.11.ffn_gate.weight - [ 4096, 14336, 1, 1], type = f32, quantizing to q3_K .. size = 224.00 MiB -> 24.06 MiB | hist:
|
| 120 |
+
[ 90/ 291] blk.11.ffn_up.weight - [ 4096, 14336, 1, 1], type = f32, quantizing to q3_K .. size = 224.00 MiB -> 24.06 MiB | hist:
|
| 121 |
+
[ 91/ 291] blk.11.ffn_norm.weight - [ 4096, 1, 1, 1], type = f32, size = 0.016 MB
|
| 122 |
+
[ 92/ 291] blk.11.attn_k.weight - [ 4096, 1024, 1, 1], type = f32, quantizing to q3_K .. size = 16.00 MiB -> 1.72 MiB | hist:
|
| 123 |
+
[ 93/ 291] blk.11.attn_output.weight - [ 4096, 4096, 1, 1], type = f32, quantizing to q4_K .. size = 64.00 MiB -> 9.00 MiB | hist:
|
| 124 |
+
[ 94/ 291] blk.11.attn_q.weight - [ 4096, 4096, 1, 1], type = f32, quantizing to q3_K .. size = 64.00 MiB -> 6.88 MiB | hist:
|
| 125 |
+
[ 95/ 291] blk.11.attn_v.weight - [ 4096, 1024, 1, 1], type = f32, quantizing to q4_K .. size = 16.00 MiB -> 2.25 MiB | hist:
|
| 126 |
+
[ 96/ 291] blk.12.ffn_gate.weight - [ 4096, 14336, 1, 1], type = f32, quantizing to q3_K .. size = 224.00 MiB -> 24.06 MiB | hist:
|
| 127 |
+
[ 97/ 291] blk.12.ffn_up.weight - [ 4096, 14336, 1, 1], type = f32, quantizing to q3_K .. size = 224.00 MiB -> 24.06 MiB | hist:
|
| 128 |
+
[ 98/ 291] blk.12.attn_k.weight - [ 4096, 1024, 1, 1], type = f32, quantizing to q3_K .. size = 16.00 MiB -> 1.72 MiB | hist:
|
| 129 |
+
[ 99/ 291] blk.12.attn_output.weight - [ 4096, 4096, 1, 1], type = f32, quantizing to q4_K .. size = 64.00 MiB -> 9.00 MiB | hist:
|
| 130 |
+
[ 100/ 291] blk.12.attn_q.weight - [ 4096, 4096, 1, 1], type = f32, quantizing to q3_K .. size = 64.00 MiB -> 6.88 MiB | hist:
|
| 131 |
+
[ 101/ 291] blk.12.attn_v.weight - [ 4096, 1024, 1, 1], type = f32, quantizing to q4_K .. size = 16.00 MiB -> 2.25 MiB | hist:
|
| 132 |
+
[ 102/ 291] blk.8.attn_norm.weight - [ 4096, 1, 1, 1], type = f32, size = 0.016 MB
|
| 133 |
+
[ 103/ 291] blk.8.ffn_down.weight - [14336, 4096, 1, 1], type = f32, quantizing to q4_K .. size = 224.00 MiB -> 31.50 MiB | hist:
|
| 134 |
+
[ 104/ 291] blk.8.ffn_gate.weight - [ 4096, 14336, 1, 1], type = f32, quantizing to q3_K .. size = 224.00 MiB -> 24.06 MiB | hist:
|
| 135 |
+
[ 105/ 291] blk.8.ffn_up.weight - [ 4096, 14336, 1, 1], type = f32, quantizing to q3_K .. size = 224.00 MiB -> 24.06 MiB | hist:
|
| 136 |
+
[ 106/ 291] blk.8.ffn_norm.weight - [ 4096, 1, 1, 1], type = f32, size = 0.016 MB
|
| 137 |
+
[ 107/ 291] blk.9.attn_norm.weight - [ 4096, 1, 1, 1], type = f32, size = 0.016 MB
|
| 138 |
+
[ 108/ 291] blk.9.ffn_down.weight - [14336, 4096, 1, 1], type = f32, quantizing to q4_K .. size = 224.00 MiB -> 31.50 MiB | hist:
|
| 139 |
+
[ 109/ 291] blk.9.ffn_gate.weight - [ 4096, 14336, 1, 1], type = f32, quantizing to q3_K .. size = 224.00 MiB -> 24.06 MiB | hist:
|
| 140 |
+
[ 110/ 291] blk.9.ffn_up.weight - [ 4096, 14336, 1, 1], type = f32, quantizing to q3_K .. size = 224.00 MiB -> 24.06 MiB | hist:
|
| 141 |
+
[ 111/ 291] blk.9.ffn_norm.weight - [ 4096, 1, 1, 1], type = f32, size = 0.016 MB
|
| 142 |
+
[ 112/ 291] blk.9.attn_k.weight - [ 4096, 1024, 1, 1], type = f32, quantizing to q3_K .. size = 16.00 MiB -> 1.72 MiB | hist:
|
| 143 |
+
[ 113/ 291] blk.9.attn_output.weight - [ 4096, 4096, 1, 1], type = f32, quantizing to q4_K .. size = 64.00 MiB -> 9.00 MiB | hist:
|
| 144 |
+
[ 114/ 291] blk.9.attn_q.weight - [ 4096, 4096, 1, 1], type = f32, quantizing to q3_K .. size = 64.00 MiB -> 6.88 MiB | hist:
|
| 145 |
+
[ 115/ 291] blk.9.attn_v.weight - [ 4096, 1024, 1, 1], type = f32, quantizing to q4_K .. size = 16.00 MiB -> 2.25 MiB | hist:
|
| 146 |
+
[ 116/ 291] blk.12.attn_norm.weight - [ 4096, 1, 1, 1], type = f32, size = 0.016 MB
|
| 147 |
+
[ 117/ 291] blk.12.ffn_down.weight - [14336, 4096, 1, 1], type = f32, quantizing to q4_K .. size = 224.00 MiB -> 31.50 MiB | hist:
|
| 148 |
+
[ 118/ 291] blk.12.ffn_norm.weight - [ 4096, 1, 1, 1], type = f32, size = 0.016 MB
|
| 149 |
+
[ 119/ 291] blk.13.attn_norm.weight - [ 4096, 1, 1, 1], type = f32, size = 0.016 MB
|
| 150 |
+
[ 120/ 291] blk.13.ffn_down.weight - [14336, 4096, 1, 1], type = f32, quantizing to q4_K .. size = 224.00 MiB -> 31.50 MiB | hist:
|
| 151 |
+
[ 121/ 291] blk.13.ffn_gate.weight - [ 4096, 14336, 1, 1], type = f32, quantizing to q3_K .. size = 224.00 MiB -> 24.06 MiB | hist:
|
| 152 |
+
[ 122/ 291] blk.13.ffn_up.weight - [ 4096, 14336, 1, 1], type = f32, quantizing to q3_K .. size = 224.00 MiB -> 24.06 MiB | hist:
|
| 153 |
+
[ 123/ 291] blk.13.ffn_norm.weight - [ 4096, 1, 1, 1], type = f32, size = 0.016 MB
|
| 154 |
+
[ 124/ 291] blk.13.attn_k.weight - [ 4096, 1024, 1, 1], type = f32, quantizing to q3_K .. size = 16.00 MiB -> 1.72 MiB | hist:
|
| 155 |
+
[ 125/ 291] blk.13.attn_output.weight - [ 4096, 4096, 1, 1], type = f32, quantizing to q4_K .. size = 64.00 MiB -> 9.00 MiB | hist:
|
| 156 |
+
[ 126/ 291] blk.13.attn_q.weight - [ 4096, 4096, 1, 1], type = f32, quantizing to q3_K .. size = 64.00 MiB -> 6.88 MiB | hist:
|
| 157 |
+
[ 127/ 291] blk.13.attn_v.weight - [ 4096, 1024, 1, 1], type = f32, quantizing to q4_K .. size = 16.00 MiB -> 2.25 MiB | hist:
|
| 158 |
+
[ 128/ 291] blk.14.attn_norm.weight - [ 4096, 1, 1, 1], type = f32, size = 0.016 MB
|
| 159 |
+
[ 129/ 291] blk.14.ffn_down.weight - [14336, 4096, 1, 1], type = f32, quantizing to q4_K .. size = 224.00 MiB -> 31.50 MiB | hist:
|
| 160 |
+
[ 130/ 291] blk.14.ffn_gate.weight - [ 4096, 14336, 1, 1], type = f32, quantizing to q3_K .. size = 224.00 MiB -> 24.06 MiB | hist:
|
| 161 |
+
[ 131/ 291] blk.14.ffn_up.weight - [ 4096, 14336, 1, 1], type = f32, quantizing to q3_K .. size = 224.00 MiB -> 24.06 MiB | hist:
|
| 162 |
+
[ 132/ 291] blk.14.ffn_norm.weight - [ 4096, 1, 1, 1], type = f32, size = 0.016 MB
|
| 163 |
+
[ 133/ 291] blk.14.attn_k.weight - [ 4096, 1024, 1, 1], type = f32, quantizing to q3_K .. size = 16.00 MiB -> 1.72 MiB | hist:
|
| 164 |
+
[ 134/ 291] blk.14.attn_output.weight - [ 4096, 4096, 1, 1], type = f32, quantizing to q4_K .. size = 64.00 MiB -> 9.00 MiB | hist:
|
| 165 |
+
[ 135/ 291] blk.14.attn_q.weight - [ 4096, 4096, 1, 1], type = f32, quantizing to q3_K .. size = 64.00 MiB -> 6.88 MiB | hist:
|
| 166 |
+
[ 136/ 291] blk.14.attn_v.weight - [ 4096, 1024, 1, 1], type = f32, quantizing to q4_K .. size = 16.00 MiB -> 2.25 MiB | hist:
|
| 167 |
+
[ 137/ 291] blk.15.attn_norm.weight - [ 4096, 1, 1, 1], type = f32, size = 0.016 MB
|
| 168 |
+
[ 138/ 291] blk.15.ffn_down.weight - [14336, 4096, 1, 1], type = f32, quantizing to q4_K .. size = 224.00 MiB -> 31.50 MiB | hist:
|
| 169 |
+
[ 139/ 291] blk.15.ffn_gate.weight - [ 4096, 14336, 1, 1], type = f32, quantizing to q3_K .. size = 224.00 MiB -> 24.06 MiB | hist:
|
| 170 |
+
[ 140/ 291] blk.15.ffn_up.weight - [ 4096, 14336, 1, 1], type = f32, quantizing to q3_K .. size = 224.00 MiB -> 24.06 MiB | hist:
|
| 171 |
+
[ 141/ 291] blk.15.ffn_norm.weight - [ 4096, 1, 1, 1], type = f32, size = 0.016 MB
|
| 172 |
+
[ 142/ 291] blk.15.attn_k.weight - [ 4096, 1024, 1, 1], type = f32, quantizing to q3_K .. size = 16.00 MiB -> 1.72 MiB | hist:
|
| 173 |
+
[ 143/ 291] blk.15.attn_output.weight - [ 4096, 4096, 1, 1], type = f32, quantizing to q4_K .. size = 64.00 MiB -> 9.00 MiB | hist:
|
| 174 |
+
[ 144/ 291] blk.15.attn_q.weight - [ 4096, 4096, 1, 1], type = f32, quantizing to q3_K .. size = 64.00 MiB -> 6.88 MiB | hist:
|
| 175 |
+
[ 145/ 291] blk.15.attn_v.weight - [ 4096, 1024, 1, 1], type = f32, quantizing to q4_K .. size = 16.00 MiB -> 2.25 MiB | hist:
|
| 176 |
+
[ 146/ 291] blk.16.attn_norm.weight - [ 4096, 1, 1, 1], type = f32, size = 0.016 MB
|
| 177 |
+
[ 147/ 291] blk.16.ffn_down.weight - [14336, 4096, 1, 1], type = f32, quantizing to q4_K .. size = 224.00 MiB -> 31.50 MiB | hist:
|
| 178 |
+
[ 148/ 291] blk.16.ffn_gate.weight - [ 4096, 14336, 1, 1], type = f32, quantizing to q3_K .. size = 224.00 MiB -> 24.06 MiB | hist:
|
| 179 |
+
[ 149/ 291] blk.16.ffn_up.weight - [ 4096, 14336, 1, 1], type = f32, quantizing to q3_K .. size = 224.00 MiB -> 24.06 MiB | hist:
|
| 180 |
+
[ 150/ 291] blk.16.ffn_norm.weight - [ 4096, 1, 1, 1], type = f32, size = 0.016 MB
|
| 181 |
+
[ 151/ 291] blk.16.attn_k.weight - [ 4096, 1024, 1, 1], type = f32, quantizing to q3_K .. size = 16.00 MiB -> 1.72 MiB | hist:
|
| 182 |
+
[ 152/ 291] blk.16.attn_output.weight - [ 4096, 4096, 1, 1], type = f32, quantizing to q4_K .. size = 64.00 MiB -> 9.00 MiB | hist:
|
| 183 |
+
[ 153/ 291] blk.16.attn_q.weight - [ 4096, 4096, 1, 1], type = f32, quantizing to q3_K .. size = 64.00 MiB -> 6.88 MiB | hist:
|
| 184 |
+
[ 154/ 291] blk.16.attn_v.weight - [ 4096, 1024, 1, 1], type = f32, quantizing to q4_K .. size = 16.00 MiB -> 2.25 MiB | hist:
|
| 185 |
+
[ 155/ 291] blk.17.attn_k.weight - [ 4096, 1024, 1, 1], type = f32, quantizing to q3_K .. size = 16.00 MiB -> 1.72 MiB | hist:
|
| 186 |
+
[ 156/ 291] blk.17.attn_output.weight - [ 4096, 4096, 1, 1], type = f32, quantizing to q4_K .. size = 64.00 MiB -> 9.00 MiB | hist:
|
| 187 |
+
[ 157/ 291] blk.17.attn_q.weight - [ 4096, 4096, 1, 1], type = f32, quantizing to q3_K .. size = 64.00 MiB -> 6.88 MiB | hist:
|
| 188 |
+
[ 158/ 291] blk.17.attn_v.weight - [ 4096, 1024, 1, 1], type = f32, quantizing to q4_K .. size = 16.00 MiB -> 2.25 MiB | hist:
|
| 189 |
+
[ 159/ 291] blk.17.attn_norm.weight - [ 4096, 1, 1, 1], type = f32, size = 0.016 MB
|
| 190 |
+
[ 160/ 291] blk.17.ffn_down.weight - [14336, 4096, 1, 1], type = f32, quantizing to q4_K .. size = 224.00 MiB -> 31.50 MiB | hist:
|
| 191 |
+
[ 161/ 291] blk.17.ffn_gate.weight - [ 4096, 14336, 1, 1], type = f32, quantizing to q3_K .. size = 224.00 MiB -> 24.06 MiB | hist:
|
| 192 |
+
[ 162/ 291] blk.17.ffn_up.weight - [ 4096, 14336, 1, 1], type = f32, quantizing to q3_K .. size = 224.00 MiB -> 24.06 MiB | hist:
|
| 193 |
+
[ 163/ 291] blk.17.ffn_norm.weight - [ 4096, 1, 1, 1], type = f32, size = 0.016 MB
|
| 194 |
+
[ 164/ 291] blk.18.attn_norm.weight - [ 4096, 1, 1, 1], type = f32, size = 0.016 MB
|
| 195 |
+
[ 165/ 291] blk.18.ffn_down.weight - [14336, 4096, 1, 1], type = f32, quantizing to q4_K .. size = 224.00 MiB -> 31.50 MiB | hist:
|
| 196 |
+
[ 166/ 291] blk.18.ffn_gate.weight - [ 4096, 14336, 1, 1], type = f32, quantizing to q3_K .. size = 224.00 MiB -> 24.06 MiB | hist:
|
| 197 |
+
[ 167/ 291] blk.18.ffn_up.weight - [ 4096, 14336, 1, 1], type = f32, quantizing to q3_K .. size = 224.00 MiB -> 24.06 MiB | hist:
|
| 198 |
+
[ 168/ 291] blk.18.ffn_norm.weight - [ 4096, 1, 1, 1], type = f32, size = 0.016 MB
|
| 199 |
+
[ 169/ 291] blk.18.attn_k.weight - [ 4096, 1024, 1, 1], type = f32, quantizing to q3_K .. size = 16.00 MiB -> 1.72 MiB | hist:
|
| 200 |
+
[ 170/ 291] blk.18.attn_output.weight - [ 4096, 4096, 1, 1], type = f32, quantizing to q4_K .. size = 64.00 MiB -> 9.00 MiB | hist:
|
| 201 |
+
[ 171/ 291] blk.18.attn_q.weight - [ 4096, 4096, 1, 1], type = f32, quantizing to q3_K .. size = 64.00 MiB -> 6.88 MiB | hist:
|
| 202 |
+
[ 172/ 291] blk.18.attn_v.weight - [ 4096, 1024, 1, 1], type = f32, quantizing to q4_K .. size = 16.00 MiB -> 2.25 MiB | hist:
|
| 203 |
+
[ 173/ 291] blk.19.attn_norm.weight - [ 4096, 1, 1, 1], type = f32, size = 0.016 MB
|
| 204 |
+
[ 174/ 291] blk.19.ffn_down.weight - [14336, 4096, 1, 1], type = f32, quantizing to q4_K .. size = 224.00 MiB -> 31.50 MiB | hist:
|
| 205 |
+
[ 175/ 291] blk.19.ffn_gate.weight - [ 4096, 14336, 1, 1], type = f32, quantizing to q3_K .. size = 224.00 MiB -> 24.06 MiB | hist:
|
| 206 |
+
[ 176/ 291] blk.19.ffn_up.weight - [ 4096, 14336, 1, 1], type = f32, quantizing to q3_K .. size = 224.00 MiB -> 24.06 MiB | hist:
|
| 207 |
+
[ 177/ 291] blk.19.ffn_norm.weight - [ 4096, 1, 1, 1], type = f32, size = 0.016 MB
|
| 208 |
+
[ 178/ 291] blk.19.attn_k.weight - [ 4096, 1024, 1, 1], type = f32, quantizing to q3_K .. size = 16.00 MiB -> 1.72 MiB | hist:
|
| 209 |
+
[ 179/ 291] blk.19.attn_output.weight - [ 4096, 4096, 1, 1], type = f32, quantizing to q4_K .. size = 64.00 MiB -> 9.00 MiB | hist:
|
| 210 |
+
[ 180/ 291] blk.19.attn_q.weight - [ 4096, 4096, 1, 1], type = f32, quantizing to q3_K .. size = 64.00 MiB -> 6.88 MiB | hist:
|
| 211 |
+
[ 181/ 291] blk.19.attn_v.weight - [ 4096, 1024, 1, 1], type = f32, quantizing to q4_K .. size = 16.00 MiB -> 2.25 MiB | hist:
|
| 212 |
+
[ 182/ 291] blk.20.attn_norm.weight - [ 4096, 1, 1, 1], type = f32, size = 0.016 MB
|
| 213 |
+
[ 183/ 291] blk.20.ffn_down.weight - [14336, 4096, 1, 1], type = f32, quantizing to q4_K .. size = 224.00 MiB -> 31.50 MiB | hist:
|
| 214 |
+
[ 184/ 291] blk.20.ffn_gate.weight - [ 4096, 14336, 1, 1], type = f32, quantizing to q3_K .. size = 224.00 MiB -> 24.06 MiB | hist:
|
| 215 |
+
[ 185/ 291] blk.20.ffn_up.weight - [ 4096, 14336, 1, 1], type = f32, quantizing to q3_K .. size = 224.00 MiB -> 24.06 MiB | hist:
|
| 216 |
+
[ 186/ 291] blk.20.ffn_norm.weight - [ 4096, 1, 1, 1], type = f32, size = 0.016 MB
|
| 217 |
+
[ 187/ 291] blk.20.attn_k.weight - [ 4096, 1024, 1, 1], type = f32, quantizing to q3_K .. size = 16.00 MiB -> 1.72 MiB | hist:
|
| 218 |
+
[ 188/ 291] blk.20.attn_output.weight - [ 4096, 4096, 1, 1], type = f32, quantizing to q4_K .. size = 64.00 MiB -> 9.00 MiB | hist:
|
| 219 |
+
[ 189/ 291] blk.20.attn_q.weight - [ 4096, 4096, 1, 1], type = f32, quantizing to q3_K .. size = 64.00 MiB -> 6.88 MiB | hist:
|
| 220 |
+
[ 190/ 291] blk.20.attn_v.weight - [ 4096, 1024, 1, 1], type = f32, quantizing to q4_K .. size = 16.00 MiB -> 2.25 MiB | hist:
|
| 221 |
+
[ 191/ 291] blk.21.ffn_gate.weight - [ 4096, 14336, 1, 1], type = f32, quantizing to q3_K .. size = 224.00 MiB -> 24.06 MiB | hist:
|
| 222 |
+
[ 192/ 291] blk.21.ffn_up.weight - [ 4096, 14336, 1, 1], type = f32, quantizing to q3_K .. size = 224.00 MiB -> 24.06 MiB | hist:
|
| 223 |
+
[ 193/ 291] blk.21.attn_k.weight - [ 4096, 1024, 1, 1], type = f32, quantizing to q3_K .. size = 16.00 MiB -> 1.72 MiB | hist:
|
| 224 |
+
[ 194/ 291] blk.21.attn_output.weight - [ 4096, 4096, 1, 1], type = f32, quantizing to q4_K .. size = 64.00 MiB -> 9.00 MiB | hist:
|
| 225 |
+
[ 195/ 291] blk.21.attn_q.weight - [ 4096, 4096, 1, 1], type = f32, quantizing to q3_K .. size = 64.00 MiB -> 6.88 MiB | hist:
|
| 226 |
+
[ 196/ 291] blk.21.attn_v.weight - [ 4096, 1024, 1, 1], type = f32, quantizing to q4_K .. size = 16.00 MiB -> 2.25 MiB | hist:
|
| 227 |
+
[ 197/ 291] blk.21.attn_norm.weight - [ 4096, 1, 1, 1], type = f32, size = 0.016 MB
|
| 228 |
+
[ 198/ 291] blk.21.ffn_down.weight - [14336, 4096, 1, 1], type = f32, quantizing to q4_K .. size = 224.00 MiB -> 31.50 MiB | hist:
|
| 229 |
+
[ 199/ 291] blk.21.ffn_norm.weight - [ 4096, 1, 1, 1], type = f32, size = 0.016 MB
|
| 230 |
+
[ 200/ 291] blk.22.attn_norm.weight - [ 4096, 1, 1, 1], type = f32, size = 0.016 MB
|
| 231 |
+
[ 201/ 291] blk.22.ffn_down.weight - [14336, 4096, 1, 1], type = f32, quantizing to q4_K .. size = 224.00 MiB -> 31.50 MiB | hist:
|
| 232 |
+
[ 202/ 291] blk.22.ffn_gate.weight - [ 4096, 14336, 1, 1], type = f32, quantizing to q3_K .. size = 224.00 MiB -> 24.06 MiB | hist:
|
| 233 |
+
[ 203/ 291] blk.22.ffn_up.weight - [ 4096, 14336, 1, 1], type = f32, quantizing to q3_K .. size = 224.00 MiB -> 24.06 MiB | hist:
|
| 234 |
+
[ 204/ 291] blk.22.ffn_norm.weight - [ 4096, 1, 1, 1], type = f32, size = 0.016 MB
|
| 235 |
+
[ 205/ 291] blk.22.attn_k.weight - [ 4096, 1024, 1, 1], type = f32, quantizing to q3_K .. size = 16.00 MiB -> 1.72 MiB | hist:
|
| 236 |
+
[ 206/ 291] blk.22.attn_output.weight - [ 4096, 4096, 1, 1], type = f32, quantizing to q4_K .. size = 64.00 MiB -> 9.00 MiB | hist:
|
| 237 |
+
[ 207/ 291] blk.22.attn_q.weight - [ 4096, 4096, 1, 1], type = f32, quantizing to q3_K .. size = 64.00 MiB -> 6.88 MiB | hist:
|
| 238 |
+
[ 208/ 291] blk.22.attn_v.weight - [ 4096, 1024, 1, 1], type = f32, quantizing to q4_K .. size = 16.00 MiB -> 2.25 MiB | hist:
|
| 239 |
+
[ 209/ 291] blk.23.attn_norm.weight - [ 4096, 1, 1, 1], type = f32, size = 0.016 MB
|
| 240 |
+
[ 210/ 291] blk.23.ffn_down.weight - [14336, 4096, 1, 1], type = f32, quantizing to q4_K .. size = 224.00 MiB -> 31.50 MiB | hist:
|
| 241 |
+
[ 211/ 291] blk.23.ffn_gate.weight - [ 4096, 14336, 1, 1], type = f32, quantizing to q3_K .. size = 224.00 MiB -> 24.06 MiB | hist:
|
| 242 |
+
[ 212/ 291] blk.23.ffn_up.weight - [ 4096, 14336, 1, 1], type = f32, quantizing to q3_K .. size = 224.00 MiB -> 24.06 MiB | hist:
|
| 243 |
+
[ 213/ 291] blk.23.ffn_norm.weight - [ 4096, 1, 1, 1], type = f32, size = 0.016 MB
|
| 244 |
+
[ 214/ 291] blk.23.attn_k.weight - [ 4096, 1024, 1, 1], type = f32, quantizing to q3_K .. size = 16.00 MiB -> 1.72 MiB | hist:
|
| 245 |
+
[ 215/ 291] blk.23.attn_output.weight - [ 4096, 4096, 1, 1], type = f32, quantizing to q4_K .. size = 64.00 MiB -> 9.00 MiB | hist:
|
| 246 |
+
[ 216/ 291] blk.23.attn_q.weight - [ 4096, 4096, 1, 1], type = f32, quantizing to q3_K .. size = 64.00 MiB -> 6.88 MiB | hist:
|
| 247 |
+
[ 217/ 291] blk.23.attn_v.weight - [ 4096, 1024, 1, 1], type = f32, quantizing to q4_K .. size = 16.00 MiB -> 2.25 MiB | hist:
|
| 248 |
+
[ 218/ 291] blk.24.attn_norm.weight - [ 4096, 1, 1, 1], type = f32, size = 0.016 MB
|
| 249 |
+
[ 219/ 291] blk.24.ffn_down.weight - [14336, 4096, 1, 1], type = f32, quantizing to q4_K .. size = 224.00 MiB -> 31.50 MiB | hist:
|
| 250 |
+
[ 220/ 291] blk.24.ffn_gate.weight - [ 4096, 14336, 1, 1], type = f32, quantizing to q3_K .. size = 224.00 MiB -> 24.06 MiB | hist:
|
| 251 |
+
[ 221/ 291] blk.24.ffn_up.weight - [ 4096, 14336, 1, 1], type = f32, quantizing to q3_K .. size = 224.00 MiB -> 24.06 MiB | hist:
|
| 252 |
+
[ 222/ 291] blk.24.ffn_norm.weight - [ 4096, 1, 1, 1], type = f32, size = 0.016 MB
|
| 253 |
+
[ 223/ 291] blk.24.attn_k.weight - [ 4096, 1024, 1, 1], type = f32, quantizing to q3_K .. size = 16.00 MiB -> 1.72 MiB | hist:
|
| 254 |
+
[ 224/ 291] blk.24.attn_output.weight - [ 4096, 4096, 1, 1], type = f32, quantizing to q4_K .. size = 64.00 MiB -> 9.00 MiB | hist:
|
| 255 |
+
[ 225/ 291] blk.24.attn_q.weight - [ 4096, 4096, 1, 1], type = f32, quantizing to q3_K .. size = 64.00 MiB -> 6.88 MiB | hist:
|
| 256 |
+
[ 226/ 291] blk.24.attn_v.weight - [ 4096, 1024, 1, 1], type = f32, quantizing to q4_K .. size = 16.00 MiB -> 2.25 MiB | hist:
|
| 257 |
+
[ 227/ 291] blk.25.attn_norm.weight - [ 4096, 1, 1, 1], type = f32, size = 0.016 MB
|
| 258 |
+
[ 228/ 291] blk.25.ffn_down.weight - [14336, 4096, 1, 1], type = f32, quantizing to q4_K .. size = 224.00 MiB -> 31.50 MiB | hist:
|
| 259 |
+
[ 229/ 291] blk.25.ffn_gate.weight - [ 4096, 14336, 1, 1], type = f32, quantizing to q3_K .. size = 224.00 MiB -> 24.06 MiB | hist:
|
| 260 |
+
[ 230/ 291] blk.25.ffn_up.weight - [ 4096, 14336, 1, 1], type = f32, quantizing to q3_K .. size = 224.00 MiB -> 24.06 MiB | hist:
|
| 261 |
+
[ 231/ 291] blk.25.ffn_norm.weight - [ 4096, 1, 1, 1], type = f32, size = 0.016 MB
|
| 262 |
+
[ 232/ 291] blk.25.attn_k.weight - [ 4096, 1024, 1, 1], type = f32, quantizing to q3_K .. size = 16.00 MiB -> 1.72 MiB | hist:
|
| 263 |
+
[ 233/ 291] blk.25.attn_output.weight - [ 4096, 4096, 1, 1], type = f32, quantizing to q4_K .. size = 64.00 MiB -> 9.00 MiB | hist:
|
| 264 |
+
[ 234/ 291] blk.25.attn_q.weight - [ 4096, 4096, 1, 1], type = f32, quantizing to q3_K .. size = 64.00 MiB -> 6.88 MiB | hist:
|
| 265 |
+
[ 235/ 291] blk.25.attn_v.weight - [ 4096, 1024, 1, 1], type = f32, quantizing to q4_K .. size = 16.00 MiB -> 2.25 MiB | hist:
|
| 266 |
+
[ 236/ 291] blk.26.attn_k.weight - [ 4096, 1024, 1, 1], type = f32, quantizing to q3_K .. size = 16.00 MiB -> 1.72 MiB | hist:
|
| 267 |
+
[ 237/ 291] blk.26.attn_output.weight - [ 4096, 4096, 1, 1], type = f32, quantizing to q4_K .. size = 64.00 MiB -> 9.00 MiB | hist:
|
| 268 |
+
[ 238/ 291] blk.26.attn_q.weight - [ 4096, 4096, 1, 1], type = f32, quantizing to q3_K .. size = 64.00 MiB -> 6.88 MiB | hist:
|
| 269 |
+
[ 239/ 291] blk.26.attn_v.weight - [ 4096, 1024, 1, 1], type = f32, quantizing to q4_K .. size = 16.00 MiB -> 2.25 MiB | hist:
|
| 270 |
+
[ 240/ 291] blk.26.attn_norm.weight - [ 4096, 1, 1, 1], type = f32, size = 0.016 MB
|
| 271 |
+
[ 241/ 291] blk.26.ffn_down.weight - [14336, 4096, 1, 1], type = f32, quantizing to q4_K .. size = 224.00 MiB -> 31.50 MiB | hist:
|
| 272 |
+
[ 242/ 291] blk.26.ffn_gate.weight - [ 4096, 14336, 1, 1], type = f32, quantizing to q3_K .. size = 224.00 MiB -> 24.06 MiB | hist:
|
| 273 |
+
[ 243/ 291] blk.26.ffn_up.weight - [ 4096, 14336, 1, 1], type = f32, quantizing to q3_K .. size = 224.00 MiB -> 24.06 MiB | hist:
|
| 274 |
+
[ 244/ 291] blk.26.ffn_norm.weight - [ 4096, 1, 1, 1], type = f32, size = 0.016 MB
|
| 275 |
+
[ 245/ 291] blk.27.attn_norm.weight - [ 4096, 1, 1, 1], type = f32, size = 0.016 MB
|
| 276 |
+
[ 246/ 291] blk.27.ffn_down.weight - [14336, 4096, 1, 1], type = f32, quantizing to q4_K .. size = 224.00 MiB -> 31.50 MiB | hist:
|
| 277 |
+
[ 247/ 291] blk.27.ffn_gate.weight - [ 4096, 14336, 1, 1], type = f32, quantizing to q3_K .. size = 224.00 MiB -> 24.06 MiB | hist:
|
| 278 |
+
[ 248/ 291] blk.27.ffn_up.weight - [ 4096, 14336, 1, 1], type = f32, quantizing to q3_K .. size = 224.00 MiB -> 24.06 MiB | hist:
|
| 279 |
+
[ 249/ 291] blk.27.ffn_norm.weight - [ 4096, 1, 1, 1], type = f32, size = 0.016 MB
|
| 280 |
+
[ 250/ 291] blk.27.attn_k.weight - [ 4096, 1024, 1, 1], type = f32, quantizing to q3_K .. size = 16.00 MiB -> 1.72 MiB | hist:
|
| 281 |
+
[ 251/ 291] blk.27.attn_output.weight - [ 4096, 4096, 1, 1], type = f32, quantizing to q4_K .. size = 64.00 MiB -> 9.00 MiB | hist:
|
| 282 |
+
[ 252/ 291] blk.27.attn_q.weight - [ 4096, 4096, 1, 1], type = f32, quantizing to q3_K .. size = 64.00 MiB -> 6.88 MiB | hist:
|
| 283 |
+
[ 253/ 291] blk.27.attn_v.weight - [ 4096, 1024, 1, 1], type = f32, quantizing to q4_K .. size = 16.00 MiB -> 2.25 MiB | hist:
|
| 284 |
+
[ 254/ 291] blk.28.attn_norm.weight - [ 4096, 1, 1, 1], type = f32, size = 0.016 MB
|
| 285 |
+
[ 255/ 291] blk.28.ffn_down.weight - [14336, 4096, 1, 1], type = f32, quantizing to q4_K .. size = 224.00 MiB -> 31.50 MiB | hist:
|
| 286 |
+
[ 256/ 291] blk.28.ffn_gate.weight - [ 4096, 14336, 1, 1], type = f32, quantizing to q3_K .. size = 224.00 MiB -> 24.06 MiB | hist:
|
| 287 |
+
[ 257/ 291] blk.28.ffn_up.weight - [ 4096, 14336, 1, 1], type = f32, quantizing to q3_K .. size = 224.00 MiB -> 24.06 MiB | hist:
|
| 288 |
+
[ 258/ 291] blk.28.ffn_norm.weight - [ 4096, 1, 1, 1], type = f32, size = 0.016 MB
|
| 289 |
+
[ 259/ 291] blk.28.attn_k.weight - [ 4096, 1024, 1, 1], type = f32, quantizing to q3_K .. size = 16.00 MiB -> 1.72 MiB | hist:
|
| 290 |
+
[ 260/ 291] blk.28.attn_output.weight - [ 4096, 4096, 1, 1], type = f32, quantizing to q4_K .. size = 64.00 MiB -> 9.00 MiB | hist:
|
| 291 |
+
[ 261/ 291] blk.28.attn_q.weight - [ 4096, 4096, 1, 1], type = f32, quantizing to q3_K .. size = 64.00 MiB -> 6.88 MiB | hist:
|
| 292 |
+
[ 262/ 291] blk.28.attn_v.weight - [ 4096, 1024, 1, 1], type = f32, quantizing to q4_K .. size = 16.00 MiB -> 2.25 MiB | hist:
|
| 293 |
+
[ 263/ 291] blk.29.attn_norm.weight - [ 4096, 1, 1, 1], type = f32, size = 0.016 MB
|
| 294 |
+
[ 264/ 291] blk.29.ffn_down.weight - [14336, 4096, 1, 1], type = f32, quantizing to q4_K .. size = 224.00 MiB -> 31.50 MiB | hist:
|
| 295 |
+
[ 265/ 291] blk.29.ffn_gate.weight - [ 4096, 14336, 1, 1], type = f32, quantizing to q3_K .. size = 224.00 MiB -> 24.06 MiB | hist:
|
| 296 |
+
[ 266/ 291] blk.29.ffn_up.weight - [ 4096, 14336, 1, 1], type = f32, quantizing to q3_K .. size = 224.00 MiB -> 24.06 MiB | hist:
|
| 297 |
+
[ 267/ 291] blk.29.ffn_norm.weight - [ 4096, 1, 1, 1], type = f32, size = 0.016 MB
|
| 298 |
+
[ 268/ 291] blk.29.attn_k.weight - [ 4096, 1024, 1, 1], type = f32, quantizing to q3_K .. size = 16.00 MiB -> 1.72 MiB | hist:
|
| 299 |
+
[ 269/ 291] blk.29.attn_output.weight - [ 4096, 4096, 1, 1], type = f32, quantizing to q4_K .. size = 64.00 MiB -> 9.00 MiB | hist:
|
| 300 |
+
[ 270/ 291] blk.29.attn_q.weight - [ 4096, 4096, 1, 1], type = f32, quantizing to q3_K .. size = 64.00 MiB -> 6.88 MiB | hist:
|
| 301 |
+
[ 271/ 291] blk.29.attn_v.weight - [ 4096, 1024, 1, 1], type = f32, quantizing to q4_K .. size = 16.00 MiB -> 2.25 MiB | hist:
|
| 302 |
+
[ 272/ 291] blk.30.ffn_gate.weight - [ 4096, 14336, 1, 1], type = f32, quantizing to q3_K .. size = 224.00 MiB -> 24.06 MiB | hist:
|
| 303 |
+
[ 273/ 291] blk.30.ffn_up.weight - [ 4096, 14336, 1, 1], type = f32, quantizing to q3_K .. size = 224.00 MiB -> 24.06 MiB | hist:
|
| 304 |
+
[ 274/ 291] blk.30.attn_k.weight - [ 4096, 1024, 1, 1], type = f32, quantizing to q3_K .. size = 16.00 MiB -> 1.72 MiB | hist:
|
| 305 |
+
[ 275/ 291] blk.30.attn_output.weight - [ 4096, 4096, 1, 1], type = f32, quantizing to q4_K .. size = 64.00 MiB -> 9.00 MiB | hist:
|
| 306 |
+
[ 276/ 291] blk.30.attn_q.weight - [ 4096, 4096, 1, 1], type = f32, quantizing to q3_K .. size = 64.00 MiB -> 6.88 MiB | hist:
|
| 307 |
+
[ 277/ 291] blk.30.attn_v.weight - [ 4096, 1024, 1, 1], type = f32, quantizing to q4_K .. size = 16.00 MiB -> 2.25 MiB | hist:
|
| 308 |
+
[ 278/ 291] output.weight - [ 4096, 32000, 1, 1], type = f32, quantizing to q6_K .. size = 500.00 MiB -> 102.54 MiB | hist:
|
| 309 |
+
[ 279/ 291] blk.30.attn_norm.weight - [ 4096, 1, 1, 1], type = f32, size = 0.016 MB
|
| 310 |
+
[ 280/ 291] blk.30.ffn_down.weight - [14336, 4096, 1, 1], type = f32, quantizing to q4_K .. size = 224.00 MiB -> 31.50 MiB | hist:
|
| 311 |
+
[ 281/ 291] blk.30.ffn_norm.weight - [ 4096, 1, 1, 1], type = f32, size = 0.016 MB
|
| 312 |
+
[ 282/ 291] blk.31.attn_norm.weight - [ 4096, 1, 1, 1], type = f32, size = 0.016 MB
|
| 313 |
+
[ 283/ 291] blk.31.ffn_down.weight - [14336, 4096, 1, 1], type = f32, quantizing to q4_K .. size = 224.00 MiB -> 31.50 MiB | hist:
|
| 314 |
+
[ 284/ 291] blk.31.ffn_gate.weight - [ 4096, 14336, 1, 1], type = f32, quantizing to q3_K .. size = 224.00 MiB -> 24.06 MiB | hist:
|
| 315 |
+
[ 285/ 291] blk.31.ffn_up.weight - [ 4096, 14336, 1, 1], type = f32, quantizing to q3_K .. size = 224.00 MiB -> 24.06 MiB | hist:
|
| 316 |
+
[ 286/ 291] blk.31.ffn_norm.weight - [ 4096, 1, 1, 1], type = f32, size = 0.016 MB
|
| 317 |
+
[ 287/ 291] blk.31.attn_k.weight - [ 4096, 1024, 1, 1], type = f32, quantizing to q3_K .. size = 16.00 MiB -> 1.72 MiB | hist:
|
| 318 |
+
[ 288/ 291] blk.31.attn_output.weight - [ 4096, 4096, 1, 1], type = f32, quantizing to q4_K .. size = 64.00 MiB -> 9.00 MiB | hist:
|
| 319 |
+
[ 289/ 291] blk.31.attn_q.weight - [ 4096, 4096, 1, 1], type = f32, quantizing to q3_K .. size = 64.00 MiB -> 6.88 MiB | hist:
|
| 320 |
+
[ 290/ 291] blk.31.attn_v.weight - [ 4096, 1024, 1, 1], type = f32, quantizing to q4_K .. size = 16.00 MiB -> 2.25 MiB | hist:
|
| 321 |
+
[ 291/ 291] output_norm.weight - [ 4096, 1, 1, 1], type = f32, size = 0.016 MB
|
| 322 |
+
llama_model_quantize_internal: model size = 27625.02 MB
|
| 323 |
+
llama_model_quantize_internal: quant size = 3355.27 MB
|
| 324 |
+
|
| 325 |
+
main: quantize time = 368785.09 ms
|
| 326 |
+
main: total time = 368785.09 ms
|
| 327 |
+
(dream) tb@IBM-PF38WZKF:~/funstreams/AI$ vi zephyr_f32_int3.txt
|
| 328 |
+
(dream) tb@IBM-PF38WZKF:~/funstreams/AI$ ./llama.cpp/quantize zephyr_f32.gguf zephyr_Q2_K.gguf Q2_K >zephyr_f32_int2.txt
|
| 329 |
+
main: build = 1798 (128de35)
|
| 330 |
+
main: built with cc (Ubuntu 9.4.0-1ubuntu1~20.04.2) 9.4.0 for x86_64-linux-gnu
|
| 331 |
+
main: quantizing 'zephyr_f32.gguf' to 'zephyr_Q2_K.gguf' as Q2_K
|
| 332 |
+
llama_model_loader: loaded meta data with 22 key-value pairs and 291 tensors from zephyr_f32.gguf (version GGUF V3 (latest))
|
| 333 |
+
llama_model_loader: Dumping metadata keys/values. Note: KV overrides do not apply in this output.
|
| 334 |
+
llama_model_loader: - kv 0: general.architecture str = llama
|
| 335 |
+
llama_model_loader: - kv 1: general.name str = .
|
| 336 |
+
llama_model_loader: - kv 2: llama.context_length u32 = 32768
|
| 337 |
+
llama_model_loader: - kv 3: llama.embedding_length u32 = 4096
|
| 338 |
+
llama_model_loader: - kv 4: llama.block_count u32 = 32
|
| 339 |
+
llama_model_loader: - kv 5: llama.feed_forward_length u32 = 14336
|
| 340 |
+
llama_model_loader: - kv 6: llama.rope.dimension_count u32 = 128
|
| 341 |
+
llama_model_loader: - kv 7: llama.attention.head_count u32 = 32
|
| 342 |
+
llama_model_loader: - kv 8: llama.attention.head_count_kv u32 = 8
|
| 343 |
+
llama_model_loader: - kv 9: llama.attention.layer_norm_rms_epsilon f32 = 0.000010
|
| 344 |
+
llama_model_loader: - kv 10: llama.rope.freq_base f32 = 10000.000000
|
| 345 |
+
llama_model_loader: - kv 11: general.file_type u32 = 0
|
| 346 |
+
llama_model_loader: - kv 12: tokenizer.ggml.model str = llama
|
| 347 |
+
llama_model_loader: - kv 13: tokenizer.ggml.tokens arr[str,32000] = ["<unk>", "<s>", "</s>", "<0x00>", "<...
|
| 348 |
+
llama_model_loader: - kv 14: tokenizer.ggml.scores arr[f32,32000] = [0.000000, 0.000000, 0.000000, 0.0000...
|
| 349 |
+
llama_model_loader: - kv 15: tokenizer.ggml.token_type arr[i32,32000] = [2, 3, 3, 6, 6, 6, 6, 6, 6, 6, 6, 6, ...
|
| 350 |
+
llama_model_loader: - kv 16: tokenizer.ggml.merges arr[str,58980] = ["▁ t", "i n", "e r", "▁ a", "h e...
|
| 351 |
+
llama_model_loader: - kv 17: tokenizer.ggml.bos_token_id u32 = 1
|
| 352 |
+
llama_model_loader: - kv 18: tokenizer.ggml.eos_token_id u32 = 2
|
| 353 |
+
llama_model_loader: - kv 19: tokenizer.ggml.unknown_token_id u32 = 0
|
| 354 |
+
llama_model_loader: - kv 20: tokenizer.ggml.padding_token_id u32 = 2
|
| 355 |
+
llama_model_loader: - kv 21: tokenizer.chat_template str = {% for message in messages %}\n{% if m...
|
| 356 |
+
llama_model_loader: - type f32: 291 tensors
|
| 357 |
+
llama_model_quantize_internal: meta size = 1671648 bytes
|
| 358 |
+
[ 1/ 291] token_embd.weight - [ 4096, 32000, 1, 1], type = f32, quantizing to q2_K .. size = 500.00 MiB -> 41.02 MiB | hist:
|
| 359 |
+
[ 2/ 291] blk.0.attn_norm.weight - [ 4096, 1, 1, 1], type = f32, size = 0.016 MB
|
| 360 |
+
[ 3/ 291] blk.0.ffn_down.weight - [14336, 4096, 1, 1], type = f32, quantizing to q3_K .. size = 224.00 MiB -> 24.06 MiB | hist:
|
| 361 |
+
[ 4/ 291] blk.0.ffn_gate.weight - [ 4096, 14336, 1, 1], type = f32, quantizing to q3_K .. size = 224.00 MiB -> 24.06 MiB | hist:
|
| 362 |
+
[ 5/ 291] blk.0.ffn_up.weight - [ 4096, 14336, 1, 1], type = f32, quantizing to q3_K .. size = 224.00 MiB -> 24.06 MiB | hist:
|
| 363 |
+
[ 6/ 291] blk.0.ffn_norm.weight - [ 4096, 1, 1, 1], type = f32, size = 0.016 MB
|
| 364 |
+
[ 7/ 291] blk.0.attn_k.weight - [ 4096, 1024, 1, 1], type = f32, quantizing to q2_K .. size = 16.00 MiB -> 1.31 MiB | hist:
|
| 365 |
+
[ 8/ 291] blk.0.attn_output.weight - [ 4096, 4096, 1, 1], type = f32, quantizing to q3_K .. size = 64.00 MiB -> 6.88 MiB | hist:
|
| 366 |
+
[ 9/ 291] blk.0.attn_q.weight - [ 4096, 4096, 1, 1], type = f32, quantizing to q2_K .. size = 64.00 MiB -> 5.25 MiB | hist:
|
| 367 |
+
[ 10/ 291] blk.0.attn_v.weight - [ 4096, 1024, 1, 1], type = f32, quantizing to q3_K .. size = 16.00 MiB -> 1.72 MiB | hist:
|
| 368 |
+
[ 11/ 291] blk.1.attn_norm.weight - [ 4096, 1, 1, 1], type = f32, size = 0.016 MB
|
| 369 |
+
[ 12/ 291] blk.1.ffn_down.weight - [14336, 4096, 1, 1], type = f32, quantizing to q3_K .. size = 224.00 MiB -> 24.06 MiB | hist:
|
| 370 |
+
[ 13/ 291] blk.1.ffn_gate.weight - [ 4096, 14336, 1, 1], type = f32, quantizing to q3_K .. size = 224.00 MiB -> 24.06 MiB | hist:
|
| 371 |
+
[ 14/ 291] blk.1.ffn_up.weight - [ 4096, 14336, 1, 1], type = f32, quantizing to q3_K .. size = 224.00 MiB -> 24.06 MiB | hist:
|
| 372 |
+
[ 15/ 291] blk.1.ffn_norm.weight - [ 4096, 1, 1, 1], type = f32, size = 0.016 MB
|
| 373 |
+
[ 16/ 291] blk.1.attn_k.weight - [ 4096, 1024, 1, 1], type = f32, quantizing to q2_K .. size = 16.00 MiB -> 1.31 MiB | hist:
|
| 374 |
+
[ 17/ 291] blk.1.attn_output.weight - [ 4096, 4096, 1, 1], type = f32, quantizing to q3_K .. size = 64.00 MiB -> 6.88 MiB | hist:
|
| 375 |
+
[ 18/ 291] blk.1.attn_q.weight - [ 4096, 4096, 1, 1], type = f32, quantizing to q2_K .. size = 64.00 MiB -> 5.25 MiB | hist:
|
| 376 |
+
[ 19/ 291] blk.1.attn_v.weight - [ 4096, 1024, 1, 1], type = f32, quantizing to q3_K .. size = 16.00 MiB -> 1.72 MiB | hist:
|
| 377 |
+
[ 20/ 291] blk.2.attn_norm.weight - [ 4096, 1, 1, 1], type = f32, size = 0.016 MB
|
| 378 |
+
[ 21/ 291] blk.2.ffn_down.weight - [14336, 4096, 1, 1], type = f32, quantizing to q3_K .. size = 224.00 MiB -> 24.06 MiB | hist:
|
| 379 |
+
[ 22/ 291] blk.2.ffn_gate.weight - [ 4096, 14336, 1, 1], type = f32, quantizing to q3_K .. size = 224.00 MiB -> 24.06 MiB | hist:
|
| 380 |
+
[ 23/ 291] blk.2.ffn_up.weight - [ 4096, 14336, 1, 1], type = f32, quantizing to q3_K .. size = 224.00 MiB -> 24.06 MiB | hist:
|
| 381 |
+
[ 24/ 291] blk.2.ffn_norm.weight - [ 4096, 1, 1, 1], type = f32, size = 0.016 MB
|
| 382 |
+
[ 25/ 291] blk.2.attn_k.weight - [ 4096, 1024, 1, 1], type = f32, quantizing to q2_K .. size = 16.00 MiB -> 1.31 MiB | hist:
|
| 383 |
+
[ 26/ 291] blk.2.attn_output.weight - [ 4096, 4096, 1, 1], type = f32, quantizing to q3_K .. size = 64.00 MiB -> 6.88 MiB | hist:
|
| 384 |
+
[ 27/ 291] blk.2.attn_q.weight - [ 4096, 4096, 1, 1], type = f32, quantizing to q2_K .. size = 64.00 MiB -> 5.25 MiB | hist:
|
| 385 |
+
[ 28/ 291] blk.2.attn_v.weight - [ 4096, 1024, 1, 1], type = f32, quantizing to q3_K .. size = 16.00 MiB -> 1.72 MiB | hist:
|
| 386 |
+
[ 29/ 291] blk.3.ffn_gate.weight - [ 4096, 14336, 1, 1], type = f32, quantizing to q3_K .. size = 224.00 MiB -> 24.06 MiB | hist:
|
| 387 |
+
[ 30/ 291] blk.3.ffn_up.weight - [ 4096, 14336, 1, 1], type = f32, quantizing to q3_K .. size = 224.00 MiB -> 24.06 MiB | hist:
|
| 388 |
+
[ 31/ 291] blk.3.attn_k.weight - [ 4096, 1024, 1, 1], type = f32, quantizing to q2_K .. size = 16.00 MiB -> 1.31 MiB | hist:
|
| 389 |
+
[ 32/ 291] blk.3.attn_output.weight - [ 4096, 4096, 1, 1], type = f32, quantizing to q3_K .. size = 64.00 MiB -> 6.88 MiB | hist:
|
| 390 |
+
[ 33/ 291] blk.3.attn_q.weight - [ 4096, 4096, 1, 1], type = f32, quantizing to q2_K .. size = 64.00 MiB -> 5.25 MiB | hist:
|
| 391 |
+
[ 34/ 291] blk.3.attn_v.weight - [ 4096, 1024, 1, 1], type = f32, quantizing to q3_K .. size = 16.00 MiB -> 1.72 MiB | hist:
|
| 392 |
+
[ 35/ 291] blk.3.attn_norm.weight - [ 4096, 1, 1, 1], type = f32, size = 0.016 MB
|
| 393 |
+
[ 36/ 291] blk.3.ffn_down.weight - [14336, 4096, 1, 1], type = f32, quantizing to q3_K .. size = 224.00 MiB -> 24.06 MiB | hist:
|
| 394 |
+
[ 37/ 291] blk.3.ffn_norm.weight - [ 4096, 1, 1, 1], type = f32, size = 0.016 MB
|
| 395 |
+
[ 38/ 291] blk.4.attn_norm.weight - [ 4096, 1, 1, 1], type = f32, size = 0.016 MB
|
| 396 |
+
[ 39/ 291] blk.4.ffn_down.weight - [14336, 4096, 1, 1], type = f32, quantizing to q3_K .. size = 224.00 MiB -> 24.06 MiB | hist:
|
| 397 |
+
[ 40/ 291] blk.4.ffn_gate.weight - [ 4096, 14336, 1, 1], type = f32, quantizing to q3_K .. size = 224.00 MiB -> 24.06 MiB | hist:
|
| 398 |
+
[ 41/ 291] blk.4.ffn_up.weight - [ 4096, 14336, 1, 1], type = f32, quantizing to q3_K .. size = 224.00 MiB -> 24.06 MiB | hist:
|
| 399 |
+
[ 42/ 291] blk.4.ffn_norm.weight - [ 4096, 1, 1, 1], type = f32, size = 0.016 MB
|
| 400 |
+
[ 43/ 291] blk.4.attn_k.weight - [ 4096, 1024, 1, 1], type = f32, quantizing to q2_K .. size = 16.00 MiB -> 1.31 MiB | hist:
|
| 401 |
+
[ 44/ 291] blk.4.attn_output.weight - [ 4096, 4096, 1, 1], type = f32, quantizing to q3_K .. size = 64.00 MiB -> 6.88 MiB | hist:
|
| 402 |
+
[ 45/ 291] blk.4.attn_q.weight - [ 4096, 4096, 1, 1], type = f32, quantizing to q2_K .. size = 64.00 MiB -> 5.25 MiB | hist:
|
| 403 |
+
[ 46/ 291] blk.4.attn_v.weight - [ 4096, 1024, 1, 1], type = f32, quantizing to q3_K .. size = 16.00 MiB -> 1.72 MiB | hist:
|
| 404 |
+
[ 47/ 291] blk.5.attn_norm.weight - [ 4096, 1, 1, 1], type = f32, size = 0.016 MB
|
| 405 |
+
[ 48/ 291] blk.5.ffn_down.weight - [14336, 4096, 1, 1], type = f32, quantizing to q3_K .. size = 224.00 MiB -> 24.06 MiB | hist:
|
| 406 |
+
[ 49/ 291] blk.5.ffn_gate.weight - [ 4096, 14336, 1, 1], type = f32, quantizing to q3_K .. size = 224.00 MiB -> 24.06 MiB | hist:
|
| 407 |
+
[ 50/ 291] blk.5.ffn_up.weight - [ 4096, 14336, 1, 1], type = f32, quantizing to q3_K .. size = 224.00 MiB -> 24.06 MiB | hist:
|
| 408 |
+
[ 51/ 291] blk.5.ffn_norm.weight - [ 4096, 1, 1, 1], type = f32, size = 0.016 MB
|
| 409 |
+
[ 52/ 291] blk.5.attn_k.weight - [ 4096, 1024, 1, 1], type = f32, quantizing to q2_K .. size = 16.00 MiB -> 1.31 MiB | hist:
|
| 410 |
+
[ 53/ 291] blk.5.attn_output.weight - [ 4096, 4096, 1, 1], type = f32, quantizing to q3_K .. size = 64.00 MiB -> 6.88 MiB | hist:
|
| 411 |
+
[ 54/ 291] blk.5.attn_q.weight - [ 4096, 4096, 1, 1], type = f32, quantizing to q2_K .. size = 64.00 MiB -> 5.25 MiB | hist:
|
| 412 |
+
[ 55/ 291] blk.5.attn_v.weight - [ 4096, 1024, 1, 1], type = f32, quantizing to q3_K .. size = 16.00 MiB -> 1.72 MiB | hist:
|
| 413 |
+
[ 56/ 291] blk.6.attn_norm.weight - [ 4096, 1, 1, 1], type = f32, size = 0.016 MB
|
| 414 |
+
[ 57/ 291] blk.6.ffn_down.weight - [14336, 4096, 1, 1], type = f32, quantizing to q3_K .. size = 224.00 MiB -> 24.06 MiB | hist:
|
| 415 |
+
[ 58/ 291] blk.6.ffn_gate.weight - [ 4096, 14336, 1, 1], type = f32, quantizing to q3_K .. size = 224.00 MiB -> 24.06 MiB | hist:
|
| 416 |
+
[ 59/ 291] blk.6.ffn_up.weight - [ 4096, 14336, 1, 1], type = f32, quantizing to q3_K .. size = 224.00 MiB -> 24.06 MiB | hist:
|
| 417 |
+
[ 60/ 291] blk.6.ffn_norm.weight - [ 4096, 1, 1, 1], type = f32, size = 0.016 MB
|
| 418 |
+
[ 61/ 291] blk.6.attn_k.weight - [ 4096, 1024, 1, 1], type = f32, quantizing to q2_K .. size = 16.00 MiB -> 1.31 MiB | hist:
|
| 419 |
+
[ 62/ 291] blk.6.attn_output.weight - [ 4096, 4096, 1, 1], type = f32, quantizing to q3_K .. size = 64.00 MiB -> 6.88 MiB | hist:
|
| 420 |
+
[ 63/ 291] blk.6.attn_q.weight - [ 4096, 4096, 1, 1], type = f32, quantizing to q2_K .. size = 64.00 MiB -> 5.25 MiB | hist:
|
| 421 |
+
[ 64/ 291] blk.6.attn_v.weight - [ 4096, 1024, 1, 1], type = f32, quantizing to q3_K .. size = 16.00 MiB -> 1.72 MiB | hist:
|
| 422 |
+
[ 65/ 291] blk.7.attn_norm.weight - [ 4096, 1, 1, 1], type = f32, size = 0.016 MB
|
| 423 |
+
[ 66/ 291] blk.7.ffn_down.weight - [14336, 4096, 1, 1], type = f32, quantizing to q3_K .. size = 224.00 MiB -> 24.06 MiB | hist:
|
| 424 |
+
[ 67/ 291] blk.7.ffn_gate.weight - [ 4096, 14336, 1, 1], type = f32, quantizing to q3_K .. size = 224.00 MiB -> 24.06 MiB | hist:
|
| 425 |
+
[ 68/ 291] blk.7.ffn_up.weight - [ 4096, 14336, 1, 1], type = f32, quantizing to q3_K .. size = 224.00 MiB -> 24.06 MiB | hist:
|
| 426 |
+
[ 69/ 291] blk.7.ffn_norm.weight - [ 4096, 1, 1, 1], type = f32, size = 0.016 MB
|
| 427 |
+
[ 70/ 291] blk.7.attn_k.weight - [ 4096, 1024, 1, 1], type = f32, quantizing to q2_K .. size = 16.00 MiB -> 1.31 MiB | hist:
|
| 428 |
+
[ 71/ 291] blk.7.attn_output.weight - [ 4096, 4096, 1, 1], type = f32, quantizing to q3_K .. size = 64.00 MiB -> 6.88 MiB | hist:
|
| 429 |
+
[ 72/ 291] blk.7.attn_q.weight - [ 4096, 4096, 1, 1], type = f32, quantizing to q2_K .. size = 64.00 MiB -> 5.25 MiB | hist:
|
| 430 |
+
[ 73/ 291] blk.7.attn_v.weight - [ 4096, 1024, 1, 1], type = f32, quantizing to q3_K .. size = 16.00 MiB -> 1.72 MiB | hist:
|
| 431 |
+
[ 74/ 291] blk.8.attn_k.weight - [ 4096, 1024, 1, 1], type = f32, quantizing to q2_K .. size = 16.00 MiB -> 1.31 MiB | hist:
|
| 432 |
+
[ 75/ 291] blk.8.attn_output.weight - [ 4096, 4096, 1, 1], type = f32, quantizing to q3_K .. size = 64.00 MiB -> 6.88 MiB | hist:
|
| 433 |
+
[ 76/ 291] blk.8.attn_q.weight - [ 4096, 4096, 1, 1], type = f32, quantizing to q2_K .. size = 64.00 MiB -> 5.25 MiB | hist:
|
| 434 |
+
[ 77/ 291] blk.8.attn_v.weight - [ 4096, 1024, 1, 1], type = f32, quantizing to q3_K .. size = 16.00 MiB -> 1.72 MiB | hist:
|
| 435 |
+
[ 78/ 291] blk.10.attn_norm.weight - [ 4096, 1, 1, 1], type = f32, size = 0.016 MB
|
| 436 |
+
[ 79/ 291] blk.10.ffn_down.weight - [14336, 4096, 1, 1], type = f32, quantizing to q3_K .. size = 224.00 MiB -> 24.06 MiB | hist:
|
| 437 |
+
[ 80/ 291] blk.10.ffn_gate.weight - [ 4096, 14336, 1, 1], type = f32, quantizing to q3_K .. size = 224.00 MiB -> 24.06 MiB | hist:
|
| 438 |
+
[ 81/ 291] blk.10.ffn_up.weight - [ 4096, 14336, 1, 1], type = f32, quantizing to q3_K .. size = 224.00 MiB -> 24.06 MiB | hist:
|
| 439 |
+
[ 82/ 291] blk.10.ffn_norm.weight - [ 4096, 1, 1, 1], type = f32, size = 0.016 MB
|
| 440 |
+
[ 83/ 291] blk.10.attn_k.weight - [ 4096, 1024, 1, 1], type = f32, quantizing to q2_K .. size = 16.00 MiB -> 1.31 MiB | hist:
|
| 441 |
+
[ 84/ 291] blk.10.attn_output.weight - [ 4096, 4096, 1, 1], type = f32, quantizing to q3_K .. size = 64.00 MiB -> 6.88 MiB | hist:
|
| 442 |
+
[ 85/ 291] blk.10.attn_q.weight - [ 4096, 4096, 1, 1], type = f32, quantizing to q2_K .. size = 64.00 MiB -> 5.25 MiB | hist:
|
| 443 |
+
[ 86/ 291] blk.10.attn_v.weight - [ 4096, 1024, 1, 1], type = f32, quantizing to q3_K .. size = 16.00 MiB -> 1.72 MiB | hist:
|
| 444 |
+
[ 87/ 291] blk.11.attn_norm.weight - [ 4096, 1, 1, 1], type = f32, size = 0.016 MB
|
| 445 |
+
[ 88/ 291] blk.11.ffn_down.weight - [14336, 4096, 1, 1], type = f32, quantizing to q3_K .. size = 224.00 MiB -> 24.06 MiB | hist:
|
| 446 |
+
[ 89/ 291] blk.11.ffn_gate.weight - [ 4096, 14336, 1, 1], type = f32, quantizing to q3_K .. size = 224.00 MiB -> 24.06 MiB | hist:
|
| 447 |
+
[ 90/ 291] blk.11.ffn_up.weight - [ 4096, 14336, 1, 1], type = f32, quantizing to q3_K .. size = 224.00 MiB -> 24.06 MiB | hist:
|
| 448 |
+
[ 91/ 291] blk.11.ffn_norm.weight - [ 4096, 1, 1, 1], type = f32, size = 0.016 MB
|
| 449 |
+
[ 92/ 291] blk.11.attn_k.weight - [ 4096, 1024, 1, 1], type = f32, quantizing to q2_K .. size = 16.00 MiB -> 1.31 MiB | hist:
|
| 450 |
+
[ 93/ 291] blk.11.attn_output.weight - [ 4096, 4096, 1, 1], type = f32, quantizing to q3_K .. size = 64.00 MiB -> 6.88 MiB | hist:
|
| 451 |
+
[ 94/ 291] blk.11.attn_q.weight - [ 4096, 4096, 1, 1], type = f32, quantizing to q2_K .. size = 64.00 MiB -> 5.25 MiB | hist:
|
| 452 |
+
[ 95/ 291] blk.11.attn_v.weight - [ 4096, 1024, 1, 1], type = f32, quantizing to q3_K .. size = 16.00 MiB -> 1.72 MiB | hist:
|
| 453 |
+
[ 96/ 291] blk.12.ffn_gate.weight - [ 4096, 14336, 1, 1], type = f32, quantizing to q3_K .. size = 224.00 MiB -> 24.06 MiB | hist:
|
| 454 |
+
[ 97/ 291] blk.12.ffn_up.weight - [ 4096, 14336, 1, 1], type = f32, quantizing to q3_K .. size = 224.00 MiB -> 24.06 MiB | hist:
|
| 455 |
+
[ 98/ 291] blk.12.attn_k.weight - [ 4096, 1024, 1, 1], type = f32, quantizing to q2_K .. size = 16.00 MiB -> 1.31 MiB | hist:
|
| 456 |
+
[ 99/ 291] blk.12.attn_output.weight - [ 4096, 4096, 1, 1], type = f32, quantizing to q3_K .. size = 64.00 MiB -> 6.88 MiB | hist:
|
| 457 |
+
[ 100/ 291] blk.12.attn_q.weight - [ 4096, 4096, 1, 1], type = f32, quantizing to q2_K .. size = 64.00 MiB -> 5.25 MiB | hist:
|
| 458 |
+
[ 101/ 291] blk.12.attn_v.weight - [ 4096, 1024, 1, 1], type = f32, quantizing to q3_K .. size = 16.00 MiB -> 1.72 MiB | hist:
|
| 459 |
+
[ 102/ 291] blk.8.attn_norm.weight - [ 4096, 1, 1, 1], type = f32, size = 0.016 MB
|
| 460 |
+
[ 103/ 291] blk.8.ffn_down.weight - [14336, 4096, 1, 1], type = f32, quantizing to q3_K .. size = 224.00 MiB -> 24.06 MiB | hist:
|
| 461 |
+
[ 104/ 291] blk.8.ffn_gate.weight - [ 4096, 14336, 1, 1], type = f32, quantizing to q3_K .. size = 224.00 MiB -> 24.06 MiB | hist:
|
| 462 |
+
[ 105/ 291] blk.8.ffn_up.weight - [ 4096, 14336, 1, 1], type = f32, quantizing to q3_K .. size = 224.00 MiB -> 24.06 MiB | hist:
|
| 463 |
+
[ 106/ 291] blk.8.ffn_norm.weight - [ 4096, 1, 1, 1], type = f32, size = 0.016 MB
|
| 464 |
+
[ 107/ 291] blk.9.attn_norm.weight - [ 4096, 1, 1, 1], type = f32, size = 0.016 MB
|
| 465 |
+
[ 108/ 291] blk.9.ffn_down.weight - [14336, 4096, 1, 1], type = f32, quantizing to q3_K .. size = 224.00 MiB -> 24.06 MiB | hist:
|
| 466 |
+
[ 109/ 291] blk.9.ffn_gate.weight - [ 4096, 14336, 1, 1], type = f32, quantizing to q3_K .. size = 224.00 MiB -> 24.06 MiB | hist:
|
| 467 |
+
[ 110/ 291] blk.9.ffn_up.weight - [ 4096, 14336, 1, 1], type = f32, quantizing to q3_K .. size = 224.00 MiB -> 24.06 MiB | hist:
|
| 468 |
+
[ 111/ 291] blk.9.ffn_norm.weight - [ 4096, 1, 1, 1], type = f32, size = 0.016 MB
|
| 469 |
+
[ 112/ 291] blk.9.attn_k.weight - [ 4096, 1024, 1, 1], type = f32, quantizing to q2_K .. size = 16.00 MiB -> 1.31 MiB | hist:
|
| 470 |
+
[ 113/ 291] blk.9.attn_output.weight - [ 4096, 4096, 1, 1], type = f32, quantizing to q3_K .. size = 64.00 MiB -> 6.88 MiB | hist:
|
| 471 |
+
[ 114/ 291] blk.9.attn_q.weight - [ 4096, 4096, 1, 1], type = f32, quantizing to q2_K .. size = 64.00 MiB -> 5.25 MiB | hist:
|
| 472 |
+
[ 115/ 291] blk.9.attn_v.weight - [ 4096, 1024, 1, 1], type = f32, quantizing to q3_K .. size = 16.00 MiB -> 1.72 MiB | hist:
|
| 473 |
+
[ 116/ 291] blk.12.attn_norm.weight - [ 4096, 1, 1, 1], type = f32, size = 0.016 MB
|
| 474 |
+
[ 117/ 291] blk.12.ffn_down.weight - [14336, 4096, 1, 1], type = f32, quantizing to q3_K .. size = 224.00 MiB -> 24.06 MiB | hist:
|
| 475 |
+
[ 118/ 291] blk.12.ffn_norm.weight - [ 4096, 1, 1, 1], type = f32, size = 0.016 MB
|
| 476 |
+
[ 119/ 291] blk.13.attn_norm.weight - [ 4096, 1, 1, 1], type = f32, size = 0.016 MB
|
| 477 |
+
[ 120/ 291] blk.13.ffn_down.weight - [14336, 4096, 1, 1], type = f32, quantizing to q3_K .. size = 224.00 MiB -> 24.06 MiB | hist:
|
| 478 |
+
[ 121/ 291] blk.13.ffn_gate.weight - [ 4096, 14336, 1, 1], type = f32, quantizing to q3_K .. size = 224.00 MiB -> 24.06 MiB | hist:
|
| 479 |
+
[ 122/ 291] blk.13.ffn_up.weight - [ 4096, 14336, 1, 1], type = f32, quantizing to q3_K .. size = 224.00 MiB -> 24.06 MiB | hist:
|
| 480 |
+
[ 123/ 291] blk.13.ffn_norm.weight - [ 4096, 1, 1, 1], type = f32, size = 0.016 MB
|
| 481 |
+
[ 124/ 291] blk.13.attn_k.weight - [ 4096, 1024, 1, 1], type = f32, quantizing to q2_K .. size = 16.00 MiB -> 1.31 MiB | hist:
|
| 482 |
+
[ 125/ 291] blk.13.attn_output.weight - [ 4096, 4096, 1, 1], type = f32, quantizing to q3_K .. size = 64.00 MiB -> 6.88 MiB | hist:
|
| 483 |
+
[ 126/ 291] blk.13.attn_q.weight - [ 4096, 4096, 1, 1], type = f32, quantizing to q2_K .. size = 64.00 MiB -> 5.25 MiB | hist:
|
| 484 |
+
[ 127/ 291] blk.13.attn_v.weight - [ 4096, 1024, 1, 1], type = f32, quantizing to q3_K .. size = 16.00 MiB -> 1.72 MiB | hist:
|
| 485 |
+
[ 128/ 291] blk.14.attn_norm.weight - [ 4096, 1, 1, 1], type = f32, size = 0.016 MB
|
| 486 |
+
[ 129/ 291] blk.14.ffn_down.weight - [14336, 4096, 1, 1], type = f32, quantizing to q3_K .. size = 224.00 MiB -> 24.06 MiB | hist:
|
| 487 |
+
[ 130/ 291] blk.14.ffn_gate.weight - [ 4096, 14336, 1, 1], type = f32, quantizing to q3_K .. size = 224.00 MiB -> 24.06 MiB | hist:
|
| 488 |
+
[ 131/ 291] blk.14.ffn_up.weight - [ 4096, 14336, 1, 1], type = f32, quantizing to q3_K .. size = 224.00 MiB -> 24.06 MiB | hist:
|
| 489 |
+
[ 132/ 291] blk.14.ffn_norm.weight - [ 4096, 1, 1, 1], type = f32, size = 0.016 MB
|
| 490 |
+
[ 133/ 291] blk.14.attn_k.weight - [ 4096, 1024, 1, 1], type = f32, quantizing to q2_K .. size = 16.00 MiB -> 1.31 MiB | hist:
|
| 491 |
+
[ 134/ 291] blk.14.attn_output.weight - [ 4096, 4096, 1, 1], type = f32, quantizing to q3_K .. size = 64.00 MiB -> 6.88 MiB | hist:
|
| 492 |
+
[ 135/ 291] blk.14.attn_q.weight - [ 4096, 4096, 1, 1], type = f32, quantizing to q2_K .. size = 64.00 MiB -> 5.25 MiB | hist:
|
| 493 |
+
[ 136/ 291] blk.14.attn_v.weight - [ 4096, 1024, 1, 1], type = f32, quantizing to q3_K .. size = 16.00 MiB -> 1.72 MiB | hist:
|
| 494 |
+
[ 137/ 291] blk.15.attn_norm.weight - [ 4096, 1, 1, 1], type = f32, size = 0.016 MB
|
| 495 |
+
[ 138/ 291] blk.15.ffn_down.weight - [14336, 4096, 1, 1], type = f32, quantizing to q3_K .. size = 224.00 MiB -> 24.06 MiB | hist:
|
| 496 |
+
[ 139/ 291] blk.15.ffn_gate.weight - [ 4096, 14336, 1, 1], type = f32, quantizing to q3_K .. size = 224.00 MiB -> 24.06 MiB | hist:
|
| 497 |
+
[ 140/ 291] blk.15.ffn_up.weight - [ 4096, 14336, 1, 1], type = f32, quantizing to q3_K .. size = 224.00 MiB -> 24.06 MiB | hist:
|
| 498 |
+
[ 141/ 291] blk.15.ffn_norm.weight - [ 4096, 1, 1, 1], type = f32, size = 0.016 MB
|
| 499 |
+
[ 142/ 291] blk.15.attn_k.weight - [ 4096, 1024, 1, 1], type = f32, quantizing to q2_K .. size = 16.00 MiB -> 1.31 MiB | hist:
|
| 500 |
+
[ 143/ 291] blk.15.attn_output.weight - [ 4096, 4096, 1, 1], type = f32, quantizing to q3_K .. size = 64.00 MiB -> 6.88 MiB | hist:
|
| 501 |
+
[ 144/ 291] blk.15.attn_q.weight - [ 4096, 4096, 1, 1], type = f32, quantizing to q2_K .. size = 64.00 MiB -> 5.25 MiB | hist:
|
| 502 |
+
[ 145/ 291] blk.15.attn_v.weight - [ 4096, 1024, 1, 1], type = f32, quantizing to q3_K .. size = 16.00 MiB -> 1.72 MiB | hist:
|
| 503 |
+
[ 146/ 291] blk.16.attn_norm.weight - [ 4096, 1, 1, 1], type = f32, size = 0.016 MB
|
| 504 |
+
[ 147/ 291] blk.16.ffn_down.weight - [14336, 4096, 1, 1], type = f32, quantizing to q3_K .. size = 224.00 MiB -> 24.06 MiB | hist:
|
| 505 |
+
[ 148/ 291] blk.16.ffn_gate.weight - [ 4096, 14336, 1, 1], type = f32, quantizing to q3_K .. size = 224.00 MiB -> 24.06 MiB | hist:
|
| 506 |
+
[ 149/ 291] blk.16.ffn_up.weight - [ 4096, 14336, 1, 1], type = f32, quantizing to q3_K .. size = 224.00 MiB -> 24.06 MiB | hist:
|
| 507 |
+
[ 150/ 291] blk.16.ffn_norm.weight - [ 4096, 1, 1, 1], type = f32, size = 0.016 MB
|
| 508 |
+
[ 151/ 291] blk.16.attn_k.weight - [ 4096, 1024, 1, 1], type = f32, quantizing to q2_K .. size = 16.00 MiB -> 1.31 MiB | hist:
|
| 509 |
+
[ 152/ 291] blk.16.attn_output.weight - [ 4096, 4096, 1, 1], type = f32, quantizing to q3_K .. size = 64.00 MiB -> 6.88 MiB | hist:
|
| 510 |
+
[ 153/ 291] blk.16.attn_q.weight - [ 4096, 4096, 1, 1], type = f32, quantizing to q2_K .. size = 64.00 MiB -> 5.25 MiB | hist:
|
| 511 |
+
[ 154/ 291] blk.16.attn_v.weight - [ 4096, 1024, 1, 1], type = f32, quantizing to q3_K .. size = 16.00 MiB -> 1.72 MiB | hist:
|
| 512 |
+
[ 155/ 291] blk.17.attn_k.weight - [ 4096, 1024, 1, 1], type = f32, quantizing to q2_K .. size = 16.00 MiB -> 1.31 MiB | hist:
|
| 513 |
+
[ 156/ 291] blk.17.attn_output.weight - [ 4096, 4096, 1, 1], type = f32, quantizing to q3_K .. size = 64.00 MiB -> 6.88 MiB | hist:
|
| 514 |
+
[ 157/ 291] blk.17.attn_q.weight - [ 4096, 4096, 1, 1], type = f32, quantizing to q2_K .. size = 64.00 MiB -> 5.25 MiB | hist:
|
| 515 |
+
[ 158/ 291] blk.17.attn_v.weight - [ 4096, 1024, 1, 1], type = f32, quantizing to q3_K .. size = 16.00 MiB -> 1.72 MiB | hist:
|
| 516 |
+
[ 159/ 291] blk.17.attn_norm.weight - [ 4096, 1, 1, 1], type = f32, size = 0.016 MB
|
| 517 |
+
[ 160/ 291] blk.17.ffn_down.weight - [14336, 4096, 1, 1], type = f32, quantizing to q3_K .. size = 224.00 MiB -> 24.06 MiB | hist:
|
| 518 |
+
[ 161/ 291] blk.17.ffn_gate.weight - [ 4096, 14336, 1, 1], type = f32, quantizing to q3_K .. size = 224.00 MiB -> 24.06 MiB | hist:
|
| 519 |
+
[ 162/ 291] blk.17.ffn_up.weight - [ 4096, 14336, 1, 1], type = f32, quantizing to q3_K .. size = 224.00 MiB -> 24.06 MiB | hist:
|
| 520 |
+
[ 163/ 291] blk.17.ffn_norm.weight - [ 4096, 1, 1, 1], type = f32, size = 0.016 MB
|
| 521 |
+
[ 164/ 291] blk.18.attn_norm.weight - [ 4096, 1, 1, 1], type = f32, size = 0.016 MB
|
| 522 |
+
[ 165/ 291] blk.18.ffn_down.weight - [14336, 4096, 1, 1], type = f32, quantizing to q3_K .. size = 224.00 MiB -> 24.06 MiB | hist:
|
| 523 |
+
[ 166/ 291] blk.18.ffn_gate.weight - [ 4096, 14336, 1, 1], type = f32, quantizing to q3_K .. size = 224.00 MiB -> 24.06 MiB | hist:
|
| 524 |
+
[ 167/ 291] blk.18.ffn_up.weight - [ 4096, 14336, 1, 1], type = f32, quantizing to q3_K .. size = 224.00 MiB -> 24.06 MiB | hist:
|
| 525 |
+
[ 168/ 291] blk.18.ffn_norm.weight - [ 4096, 1, 1, 1], type = f32, size = 0.016 MB
|
| 526 |
+
[ 169/ 291] blk.18.attn_k.weight - [ 4096, 1024, 1, 1], type = f32, quantizing to q2_K .. size = 16.00 MiB -> 1.31 MiB | hist:
|
| 527 |
+
[ 170/ 291] blk.18.attn_output.weight - [ 4096, 4096, 1, 1], type = f32, quantizing to q3_K .. size = 64.00 MiB -> 6.88 MiB | hist:
|
| 528 |
+
[ 171/ 291] blk.18.attn_q.weight - [ 4096, 4096, 1, 1], type = f32, quantizing to q2_K .. size = 64.00 MiB -> 5.25 MiB | hist:
|
| 529 |
+
[ 172/ 291] blk.18.attn_v.weight - [ 4096, 1024, 1, 1], type = f32, quantizing to q3_K .. size = 16.00 MiB -> 1.72 MiB | hist:
|
| 530 |
+
[ 173/ 291] blk.19.attn_norm.weight - [ 4096, 1, 1, 1], type = f32, size = 0.016 MB
|
| 531 |
+
[ 174/ 291] blk.19.ffn_down.weight - [14336, 4096, 1, 1], type = f32, quantizing to q3_K .. size = 224.00 MiB -> 24.06 MiB | hist:
|
| 532 |
+
[ 175/ 291] blk.19.ffn_gate.weight - [ 4096, 14336, 1, 1], type = f32, quantizing to q3_K .. size = 224.00 MiB -> 24.06 MiB | hist:
|
| 533 |
+
[ 176/ 291] blk.19.ffn_up.weight - [ 4096, 14336, 1, 1], type = f32, quantizing to q3_K .. size = 224.00 MiB -> 24.06 MiB | hist:
|
| 534 |
+
[ 177/ 291] blk.19.ffn_norm.weight - [ 4096, 1, 1, 1], type = f32, size = 0.016 MB
|
| 535 |
+
[ 178/ 291] blk.19.attn_k.weight - [ 4096, 1024, 1, 1], type = f32, quantizing to q2_K .. size = 16.00 MiB -> 1.31 MiB | hist:
|
| 536 |
+
[ 179/ 291] blk.19.attn_output.weight - [ 4096, 4096, 1, 1], type = f32, quantizing to q3_K .. size = 64.00 MiB -> 6.88 MiB | hist:
|
| 537 |
+
[ 180/ 291] blk.19.attn_q.weight - [ 4096, 4096, 1, 1], type = f32, quantizing to q2_K .. size = 64.00 MiB -> 5.25 MiB | hist:
|
| 538 |
+
[ 181/ 291] blk.19.attn_v.weight - [ 4096, 1024, 1, 1], type = f32, quantizing to q3_K .. size = 16.00 MiB -> 1.72 MiB | hist:
|
| 539 |
+
[ 182/ 291] blk.20.attn_norm.weight - [ 4096, 1, 1, 1], type = f32, size = 0.016 MB
|
| 540 |
+
[ 183/ 291] blk.20.ffn_down.weight - [14336, 4096, 1, 1], type = f32, quantizing to q3_K .. size = 224.00 MiB -> 24.06 MiB | hist:
|
| 541 |
+
[ 184/ 291] blk.20.ffn_gate.weight - [ 4096, 14336, 1, 1], type = f32, quantizing to q3_K .. size = 224.00 MiB -> 24.06 MiB | hist:
|
| 542 |
+
[ 185/ 291] blk.20.ffn_up.weight - [ 4096, 14336, 1, 1], type = f32, quantizing to q3_K .. size = 224.00 MiB -> 24.06 MiB | hist:
|
| 543 |
+
[ 186/ 291] blk.20.ffn_norm.weight - [ 4096, 1, 1, 1], type = f32, size = 0.016 MB
|
| 544 |
+
[ 187/ 291] blk.20.attn_k.weight - [ 4096, 1024, 1, 1], type = f32, quantizing to q2_K .. size = 16.00 MiB -> 1.31 MiB | hist:
|
| 545 |
+
[ 188/ 291] blk.20.attn_output.weight - [ 4096, 4096, 1, 1], type = f32, quantizing to q3_K .. size = 64.00 MiB -> 6.88 MiB | hist:
|
| 546 |
+
[ 189/ 291] blk.20.attn_q.weight - [ 4096, 4096, 1, 1], type = f32, quantizing to q2_K .. size = 64.00 MiB -> 5.25 MiB | hist:
|
| 547 |
+
[ 190/ 291] blk.20.attn_v.weight - [ 4096, 1024, 1, 1], type = f32, quantizing to q3_K .. size = 16.00 MiB -> 1.72 MiB | hist:
|
| 548 |
+
[ 191/ 291] blk.21.ffn_gate.weight - [ 4096, 14336, 1, 1], type = f32, quantizing to q3_K .. size = 224.00 MiB -> 24.06 MiB | hist:
|
| 549 |
+
[ 192/ 291] blk.21.ffn_up.weight - [ 4096, 14336, 1, 1], type = f32, quantizing to q3_K .. size = 224.00 MiB -> 24.06 MiB | hist:
|
| 550 |
+
[ 193/ 291] blk.21.attn_k.weight - [ 4096, 1024, 1, 1], type = f32, quantizing to q2_K .. size = 16.00 MiB -> 1.31 MiB | hist:
|
| 551 |
+
[ 194/ 291] blk.21.attn_output.weight - [ 4096, 4096, 1, 1], type = f32, quantizing to q3_K .. size = 64.00 MiB -> 6.88 MiB | hist:
|
| 552 |
+
[ 195/ 291] blk.21.attn_q.weight - [ 4096, 4096, 1, 1], type = f32, quantizing to q2_K .. size = 64.00 MiB -> 5.25 MiB | hist:
|
| 553 |
+
[ 196/ 291] blk.21.attn_v.weight - [ 4096, 1024, 1, 1], type = f32, quantizing to q3_K .. size = 16.00 MiB -> 1.72 MiB | hist:
|
| 554 |
+
[ 197/ 291] blk.21.attn_norm.weight - [ 4096, 1, 1, 1], type = f32, size = 0.016 MB
|
| 555 |
+
[ 198/ 291] blk.21.ffn_down.weight - [14336, 4096, 1, 1], type = f32, quantizing to q3_K .. size = 224.00 MiB -> 24.06 MiB | hist:
|
| 556 |
+
[ 199/ 291] blk.21.ffn_norm.weight - [ 4096, 1, 1, 1], type = f32, size = 0.016 MB
|
| 557 |
+
[ 200/ 291] blk.22.attn_norm.weight - [ 4096, 1, 1, 1], type = f32, size = 0.016 MB
|
| 558 |
+
[ 201/ 291] blk.22.ffn_down.weight - [14336, 4096, 1, 1], type = f32, quantizing to q3_K .. size = 224.00 MiB -> 24.06 MiB | hist:
|
| 559 |
+
[ 202/ 291] blk.22.ffn_gate.weight - [ 4096, 14336, 1, 1], type = f32, quantizing to q3_K .. size = 224.00 MiB -> 24.06 MiB | hist:
|
| 560 |
+
[ 203/ 291] blk.22.ffn_up.weight - [ 4096, 14336, 1, 1], type = f32, quantizing to q3_K .. size = 224.00 MiB -> 24.06 MiB | hist:
|
| 561 |
+
[ 204/ 291] blk.22.ffn_norm.weight - [ 4096, 1, 1, 1], type = f32, size = 0.016 MB
|
| 562 |
+
[ 205/ 291] blk.22.attn_k.weight - [ 4096, 1024, 1, 1], type = f32, quantizing to q2_K .. size = 16.00 MiB -> 1.31 MiB | hist:
|
| 563 |
+
[ 206/ 291] blk.22.attn_output.weight - [ 4096, 4096, 1, 1], type = f32, quantizing to q3_K .. size = 64.00 MiB -> 6.88 MiB | hist:
|
| 564 |
+
[ 207/ 291] blk.22.attn_q.weight - [ 4096, 4096, 1, 1], type = f32, quantizing to q2_K .. size = 64.00 MiB -> 5.25 MiB | hist:
|
| 565 |
+
[ 208/ 291] blk.22.attn_v.weight - [ 4096, 1024, 1, 1], type = f32, quantizing to q3_K .. size = 16.00 MiB -> 1.72 MiB | hist:
|
| 566 |
+
[ 209/ 291] blk.23.attn_norm.weight - [ 4096, 1, 1, 1], type = f32, size = 0.016 MB
|
| 567 |
+
[ 210/ 291] blk.23.ffn_down.weight - [14336, 4096, 1, 1], type = f32, quantizing to q3_K .. size = 224.00 MiB -> 24.06 MiB | hist:
|
| 568 |
+
[ 211/ 291] blk.23.ffn_gate.weight - [ 4096, 14336, 1, 1], type = f32, quantizing to q3_K .. size = 224.00 MiB -> 24.06 MiB | hist:
|
| 569 |
+
[ 212/ 291] blk.23.ffn_up.weight - [ 4096, 14336, 1, 1], type = f32, quantizing to q3_K .. size = 224.00 MiB -> 24.06 MiB | hist:
|
| 570 |
+
[ 213/ 291] blk.23.ffn_norm.weight - [ 4096, 1, 1, 1], type = f32, size = 0.016 MB
|
| 571 |
+
[ 214/ 291] blk.23.attn_k.weight - [ 4096, 1024, 1, 1], type = f32, quantizing to q2_K .. size = 16.00 MiB -> 1.31 MiB | hist:
|
| 572 |
+
[ 215/ 291] blk.23.attn_output.weight - [ 4096, 4096, 1, 1], type = f32, quantizing to q3_K .. size = 64.00 MiB -> 6.88 MiB | hist:
|
| 573 |
+
[ 216/ 291] blk.23.attn_q.weight - [ 4096, 4096, 1, 1], type = f32, quantizing to q2_K .. size = 64.00 MiB -> 5.25 MiB | hist:
|
| 574 |
+
[ 217/ 291] blk.23.attn_v.weight - [ 4096, 1024, 1, 1], type = f32, quantizing to q3_K .. size = 16.00 MiB -> 1.72 MiB | hist:
|
| 575 |
+
[ 218/ 291] blk.24.attn_norm.weight - [ 4096, 1, 1, 1], type = f32, size = 0.016 MB
|
| 576 |
+
[ 219/ 291] blk.24.ffn_down.weight - [14336, 4096, 1, 1], type = f32, quantizing to q3_K .. size = 224.00 MiB -> 24.06 MiB | hist:
|
| 577 |
+
[ 220/ 291] blk.24.ffn_gate.weight - [ 4096, 14336, 1, 1], type = f32, quantizing to q3_K .. size = 224.00 MiB -> 24.06 MiB | hist:
|
| 578 |
+
[ 221/ 291] blk.24.ffn_up.weight - [ 4096, 14336, 1, 1], type = f32, quantizing to q3_K .. size = 224.00 MiB -> 24.06 MiB | hist:
|
| 579 |
+
[ 222/ 291] blk.24.ffn_norm.weight - [ 4096, 1, 1, 1], type = f32, size = 0.016 MB
|
| 580 |
+
[ 223/ 291] blk.24.attn_k.weight - [ 4096, 1024, 1, 1], type = f32, quantizing to q2_K .. size = 16.00 MiB -> 1.31 MiB | hist:
|
| 581 |
+
[ 224/ 291] blk.24.attn_output.weight - [ 4096, 4096, 1, 1], type = f32, quantizing to q3_K .. size = 64.00 MiB -> 6.88 MiB | hist:
|
| 582 |
+
[ 225/ 291] blk.24.attn_q.weight - [ 4096, 4096, 1, 1], type = f32, quantizing to q2_K .. size = 64.00 MiB -> 5.25 MiB | hist:
|
| 583 |
+
[ 226/ 291] blk.24.attn_v.weight - [ 4096, 1024, 1, 1], type = f32, quantizing to q3_K .. size = 16.00 MiB -> 1.72 MiB | hist:
|
| 584 |
+
[ 227/ 291] blk.25.attn_norm.weight - [ 4096, 1, 1, 1], type = f32, size = 0.016 MB
|
| 585 |
+
[ 228/ 291] blk.25.ffn_down.weight - [14336, 4096, 1, 1], type = f32, quantizing to q3_K .. size = 224.00 MiB -> 24.06 MiB | hist:
|
| 586 |
+
[ 229/ 291] blk.25.ffn_gate.weight - [ 4096, 14336, 1, 1], type = f32, quantizing to q3_K .. size = 224.00 MiB -> 24.06 MiB | hist:
|
| 587 |
+
[ 230/ 291] blk.25.ffn_up.weight - [ 4096, 14336, 1, 1], type = f32, quantizing to q3_K .. size = 224.00 MiB -> 24.06 MiB | hist:
|
| 588 |
+
[ 231/ 291] blk.25.ffn_norm.weight - [ 4096, 1, 1, 1], type = f32, size = 0.016 MB
|
| 589 |
+
[ 232/ 291] blk.25.attn_k.weight - [ 4096, 1024, 1, 1], type = f32, quantizing to q2_K .. size = 16.00 MiB -> 1.31 MiB | hist:
|
| 590 |
+
[ 233/ 291] blk.25.attn_output.weight - [ 4096, 4096, 1, 1], type = f32, quantizing to q3_K .. size = 64.00 MiB -> 6.88 MiB | hist:
|
| 591 |
+
[ 234/ 291] blk.25.attn_q.weight - [ 4096, 4096, 1, 1], type = f32, quantizing to q2_K .. size = 64.00 MiB -> 5.25 MiB | hist:
|
| 592 |
+
[ 235/ 291] blk.25.attn_v.weight - [ 4096, 1024, 1, 1], type = f32, quantizing to q3_K .. size = 16.00 MiB -> 1.72 MiB | hist:
|
| 593 |
+
[ 236/ 291] blk.26.attn_k.weight - [ 4096, 1024, 1, 1], type = f32, quantizing to q2_K .. size = 16.00 MiB -> 1.31 MiB | hist:
|
| 594 |
+
[ 237/ 291] blk.26.attn_output.weight - [ 4096, 4096, 1, 1], type = f32, quantizing to q3_K .. size = 64.00 MiB -> 6.88 MiB | hist:
|
| 595 |
+
[ 238/ 291] blk.26.attn_q.weight - [ 4096, 4096, 1, 1], type = f32, quantizing to q2_K .. size = 64.00 MiB -> 5.25 MiB | hist:
|
| 596 |
+
[ 239/ 291] blk.26.attn_v.weight - [ 4096, 1024, 1, 1], type = f32, quantizing to q3_K .. size = 16.00 MiB -> 1.72 MiB | hist:
|
| 597 |
+
[ 240/ 291] blk.26.attn_norm.weight - [ 4096, 1, 1, 1], type = f32, size = 0.016 MB
|
| 598 |
+
[ 241/ 291] blk.26.ffn_down.weight - [14336, 4096, 1, 1], type = f32, quantizing to q3_K .. size = 224.00 MiB -> 24.06 MiB | hist:
|
| 599 |
+
[ 242/ 291] blk.26.ffn_gate.weight - [ 4096, 14336, 1, 1], type = f32, quantizing to q3_K .. size = 224.00 MiB -> 24.06 MiB | hist:
|
| 600 |
+
[ 243/ 291] blk.26.ffn_up.weight - [ 4096, 14336, 1, 1], type = f32, quantizing to q3_K .. size = 224.00 MiB -> 24.06 MiB | hist:
|
| 601 |
+
[ 244/ 291] blk.26.ffn_norm.weight - [ 4096, 1, 1, 1], type = f32, size = 0.016 MB
|
| 602 |
+
[ 245/ 291] blk.27.attn_norm.weight - [ 4096, 1, 1, 1], type = f32, size = 0.016 MB
|
| 603 |
+
[ 246/ 291] blk.27.ffn_down.weight - [14336, 4096, 1, 1], type = f32, quantizing to q3_K .. size = 224.00 MiB -> 24.06 MiB | hist:
|
| 604 |
+
[ 247/ 291] blk.27.ffn_gate.weight - [ 4096, 14336, 1, 1], type = f32, quantizing to q3_K .. size = 224.00 MiB -> 24.06 MiB | hist:
|
| 605 |
+
[ 248/ 291] blk.27.ffn_up.weight - [ 4096, 14336, 1, 1], type = f32, quantizing to q3_K .. size = 224.00 MiB -> 24.06 MiB | hist:
|
| 606 |
+
[ 249/ 291] blk.27.ffn_norm.weight - [ 4096, 1, 1, 1], type = f32, size = 0.016 MB
|
| 607 |
+
[ 250/ 291] blk.27.attn_k.weight - [ 4096, 1024, 1, 1], type = f32, quantizing to q2_K .. size = 16.00 MiB -> 1.31 MiB | hist:
|
| 608 |
+
[ 251/ 291] blk.27.attn_output.weight - [ 4096, 4096, 1, 1], type = f32, quantizing to q3_K .. size = 64.00 MiB -> 6.88 MiB | hist:
|
| 609 |
+
[ 252/ 291] blk.27.attn_q.weight - [ 4096, 4096, 1, 1], type = f32, quantizing to q2_K .. size = 64.00 MiB -> 5.25 MiB | hist:
|
| 610 |
+
[ 253/ 291] blk.27.attn_v.weight - [ 4096, 1024, 1, 1], type = f32, quantizing to q3_K .. size = 16.00 MiB -> 1.72 MiB | hist:
|
| 611 |
+
[ 254/ 291] blk.28.attn_norm.weight - [ 4096, 1, 1, 1], type = f32, size = 0.016 MB
|
| 612 |
+
[ 255/ 291] blk.28.ffn_down.weight - [14336, 4096, 1, 1], type = f32, quantizing to q3_K .. size = 224.00 MiB -> 24.06 MiB | hist:
|
| 613 |
+
[ 256/ 291] blk.28.ffn_gate.weight - [ 4096, 14336, 1, 1], type = f32, quantizing to q3_K .. size = 224.00 MiB -> 24.06 MiB | hist:
|
| 614 |
+
[ 257/ 291] blk.28.ffn_up.weight - [ 4096, 14336, 1, 1], type = f32, quantizing to q3_K .. size = 224.00 MiB -> 24.06 MiB | hist:
|
| 615 |
+
[ 258/ 291] blk.28.ffn_norm.weight - [ 4096, 1, 1, 1], type = f32, size = 0.016 MB
|
| 616 |
+
[ 259/ 291] blk.28.attn_k.weight - [ 4096, 1024, 1, 1], type = f32, quantizing to q2_K .. size = 16.00 MiB -> 1.31 MiB | hist:
|
| 617 |
+
[ 260/ 291] blk.28.attn_output.weight - [ 4096, 4096, 1, 1], type = f32, quantizing to q3_K .. size = 64.00 MiB -> 6.88 MiB | hist:
|
| 618 |
+
[ 261/ 291] blk.28.attn_q.weight - [ 4096, 4096, 1, 1], type = f32, quantizing to q2_K .. size = 64.00 MiB -> 5.25 MiB | hist:
|
| 619 |
+
[ 262/ 291] blk.28.attn_v.weight - [ 4096, 1024, 1, 1], type = f32, quantizing to q3_K .. size = 16.00 MiB -> 1.72 MiB | hist:
|
| 620 |
+
[ 263/ 291] blk.29.attn_norm.weight - [ 4096, 1, 1, 1], type = f32, size = 0.016 MB
|
| 621 |
+
[ 264/ 291] blk.29.ffn_down.weight - [14336, 4096, 1, 1], type = f32, quantizing to q3_K .. size = 224.00 MiB -> 24.06 MiB | hist:
|
| 622 |
+
[ 265/ 291] blk.29.ffn_gate.weight - [ 4096, 14336, 1, 1], type = f32, quantizing to q3_K .. size = 224.00 MiB -> 24.06 MiB | hist:
|
| 623 |
+
[ 266/ 291] blk.29.ffn_up.weight - [ 4096, 14336, 1, 1], type = f32, quantizing to q3_K .. size = 224.00 MiB -> 24.06 MiB | hist:
|
| 624 |
+
[ 267/ 291] blk.29.ffn_norm.weight - [ 4096, 1, 1, 1], type = f32, size = 0.016 MB
|
| 625 |
+
[ 268/ 291] blk.29.attn_k.weight - [ 4096, 1024, 1, 1], type = f32, quantizing to q2_K .. size = 16.00 MiB -> 1.31 MiB | hist:
|
| 626 |
+
[ 269/ 291] blk.29.attn_output.weight - [ 4096, 4096, 1, 1], type = f32, quantizing to q3_K .. size = 64.00 MiB -> 6.88 MiB | hist:
|
| 627 |
+
[ 270/ 291] blk.29.attn_q.weight - [ 4096, 4096, 1, 1], type = f32, quantizing to q2_K .. size = 64.00 MiB -> 5.25 MiB | hist:
|
| 628 |
+
[ 271/ 291] blk.29.attn_v.weight - [ 4096, 1024, 1, 1], type = f32, quantizing to q3_K .. size = 16.00 MiB -> 1.72 MiB | hist:
|
| 629 |
+
[ 272/ 291] blk.30.ffn_gate.weight - [ 4096, 14336, 1, 1], type = f32, quantizing to q3_K .. size = 224.00 MiB -> 24.06 MiB | hist:
|
| 630 |
+
[ 273/ 291] blk.30.ffn_up.weight - [ 4096, 14336, 1, 1], type = f32, quantizing to q3_K .. size = 224.00 MiB -> 24.06 MiB | hist:
|
| 631 |
+
[ 274/ 291] blk.30.attn_k.weight - [ 4096, 1024, 1, 1], type = f32, quantizing to q2_K .. size = 16.00 MiB -> 1.31 MiB | hist:
|
| 632 |
+
[ 275/ 291] blk.30.attn_output.weight - [ 4096, 4096, 1, 1], type = f32, quantizing to q3_K .. size = 64.00 MiB -> 6.88 MiB | hist:
|
| 633 |
+
[ 276/ 291] blk.30.attn_q.weight - [ 4096, 4096, 1, 1], type = f32, quantizing to q2_K .. size = 64.00 MiB -> 5.25 MiB | hist:
|
| 634 |
+
[ 277/ 291] blk.30.attn_v.weight - [ 4096, 1024, 1, 1], type = f32, quantizing to q3_K .. size = 16.00 MiB -> 1.72 MiB | hist:
|
| 635 |
+
[ 278/ 291] output.weight - [ 4096, 32000, 1, 1], type = f32, quantizing to q6_K .. size = 500.00 MiB -> 102.54 MiB | hist:
|
| 636 |
+
[ 279/ 291] blk.30.attn_norm.weight - [ 4096, 1, 1, 1], type = f32, size = 0.016 MB
|
| 637 |
+
[ 280/ 291] blk.30.ffn_down.weight - [14336, 4096, 1, 1], type = f32, quantizing to q3_K .. size = 224.00 MiB -> 24.06 MiB | hist:
|
| 638 |
+
[ 281/ 291] blk.30.ffn_norm.weight - [ 4096, 1, 1, 1], type = f32, size = 0.016 MB
|
| 639 |
+
[ 282/ 291] blk.31.attn_norm.weight - [ 4096, 1, 1, 1], type = f32, size = 0.016 MB
|
| 640 |
+
[ 283/ 291] blk.31.ffn_down.weight - [14336, 4096, 1, 1], type = f32, quantizing to q3_K .. size = 224.00 MiB -> 24.06 MiB | hist:
|
| 641 |
+
[ 284/ 291] blk.31.ffn_gate.weight - [ 4096, 14336, 1, 1], type = f32, quantizing to q3_K .. size = 224.00 MiB -> 24.06 MiB | hist:
|
| 642 |
+
[ 285/ 291] blk.31.ffn_up.weight - [ 4096, 14336, 1, 1], type = f32, quantizing to q3_K .. size = 224.00 MiB -> 24.06 MiB | hist:
|
| 643 |
+
[ 286/ 291] blk.31.ffn_norm.weight - [ 4096, 1, 1, 1], type = f32, size = 0.016 MB
|
| 644 |
+
[ 287/ 291] blk.31.attn_k.weight - [ 4096, 1024, 1, 1], type = f32, quantizing to q2_K .. size = 16.00 MiB -> 1.31 MiB | hist:
|
| 645 |
+
[ 288/ 291] blk.31.attn_output.weight - [ 4096, 4096, 1, 1], type = f32, quantizing to q3_K .. size = 64.00 MiB -> 6.88 MiB | hist:
|
| 646 |
+
[ 289/ 291] blk.31.attn_q.weight - [ 4096, 4096, 1, 1], type = f32, quantizing to q2_K .. size = 64.00 MiB -> 5.25 MiB | hist:
|
| 647 |
+
[ 290/ 291] blk.31.attn_v.weight - [ 4096, 1024, 1, 1], type = f32, quantizing to q3_K .. size = 16.00 MiB -> 1.72 MiB | hist:
|
| 648 |
+
[ 291/ 291] output_norm.weight - [ 4096, 1, 1, 1], type = f32, size = 0.016 MB
|
| 649 |
+
llama_model_quantize_internal: model size = 27625.02 MB
|
| 650 |
+
llama_model_quantize_internal: quant size = 2939.57 MB
|
zephyr_f32_int3.txt
ADDED
|
@@ -0,0 +1,326 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
(dream) tb@IBM-PF38WZKF:~/funstreams/AI$ ./llama.cpp/quantize zephyr_f32.gguf Q3_K
|
| 2 |
+
main: build = 1798 (128de35)
|
| 3 |
+
main: built with cc (Ubuntu 9.4.0-1ubuntu1~20.04.2) 9.4.0 for x86_64-linux-gnu
|
| 4 |
+
main: quantizing 'zephyr_f32.gguf' to 'ggml-model-Q3_K.gguf' as Q3_K
|
| 5 |
+
llama_model_loader: loaded meta data with 22 key-value pairs and 291 tensors from zephyr_f32.gguf (version GGUF V3 (latest))
|
| 6 |
+
llama_model_loader: Dumping metadata keys/values. Note: KV overrides do not apply in this output.
|
| 7 |
+
llama_model_loader: - kv 0: general.architecture str = llama
|
| 8 |
+
llama_model_loader: - kv 1: general.name str = .
|
| 9 |
+
llama_model_loader: - kv 2: llama.context_length u32 = 32768
|
| 10 |
+
llama_model_loader: - kv 3: llama.embedding_length u32 = 4096
|
| 11 |
+
llama_model_loader: - kv 4: llama.block_count u32 = 32
|
| 12 |
+
llama_model_loader: - kv 5: llama.feed_forward_length u32 = 14336
|
| 13 |
+
llama_model_loader: - kv 6: llama.rope.dimension_count u32 = 128
|
| 14 |
+
llama_model_loader: - kv 7: llama.attention.head_count u32 = 32
|
| 15 |
+
llama_model_loader: - kv 8: llama.attention.head_count_kv u32 = 8
|
| 16 |
+
llama_model_loader: - kv 9: llama.attention.layer_norm_rms_epsilon f32 = 0.000010
|
| 17 |
+
llama_model_loader: - kv 10: llama.rope.freq_base f32 = 10000.000000
|
| 18 |
+
llama_model_loader: - kv 11: general.file_type u32 = 0
|
| 19 |
+
llama_model_loader: - kv 12: tokenizer.ggml.model str = llama
|
| 20 |
+
llama_model_loader: - kv 13: tokenizer.ggml.tokens arr[str,32000] = ["<unk>", "<s>", "</s>", "<0x00>", "<...
|
| 21 |
+
llama_model_loader: - kv 14: tokenizer.ggml.scores arr[f32,32000] = [0.000000, 0.000000, 0.000000, 0.0000...
|
| 22 |
+
llama_model_loader: - kv 15: tokenizer.ggml.token_type arr[i32,32000] = [2, 3, 3, 6, 6, 6, 6, 6, 6, 6, 6, 6, ...
|
| 23 |
+
llama_model_loader: - kv 16: tokenizer.ggml.merges arr[str,58980] = ["▁ t", "i n", "e r", "▁ a", "h e...
|
| 24 |
+
llama_model_loader: - kv 17: tokenizer.ggml.bos_token_id u32 = 1
|
| 25 |
+
llama_model_loader: - kv 18: tokenizer.ggml.eos_token_id u32 = 2
|
| 26 |
+
llama_model_loader: - kv 19: tokenizer.ggml.unknown_token_id u32 = 0
|
| 27 |
+
llama_model_loader: - kv 20: tokenizer.ggml.padding_token_id u32 = 2
|
| 28 |
+
llama_model_loader: - kv 21: tokenizer.chat_template str = {% for message in messages %}\n{% if m...
|
| 29 |
+
llama_model_loader: - type f32: 291 tensors
|
| 30 |
+
llama_model_quantize_internal: meta size = 1671648 bytes
|
| 31 |
+
[ 1/ 291] token_embd.weight - [ 4096, 32000, 1, 1], type = f32, quantizing to q3_K .. size = 500.00 MiB -> 53.71 MiB | hist:
|
| 32 |
+
[ 2/ 291] blk.0.attn_norm.weight - [ 4096, 1, 1, 1], type = f32, size = 0.016 MB
|
| 33 |
+
[ 3/ 291] blk.0.ffn_down.weight - [14336, 4096, 1, 1], type = f32, quantizing to q5_K .. size = 224.00 MiB -> 38.50 MiB | hist:
|
| 34 |
+
[ 4/ 291] blk.0.ffn_gate.weight - [ 4096, 14336, 1, 1], type = f32, quantizing to q3_K .. size = 224.00 MiB -> 24.06 MiB | hist:
|
| 35 |
+
[ 5/ 291] blk.0.ffn_up.weight - [ 4096, 14336, 1, 1], type = f32, quantizing to q3_K .. size = 224.00 MiB -> 24.06 MiB | hist:
|
| 36 |
+
[ 6/ 291] blk.0.ffn_norm.weight - [ 4096, 1, 1, 1], type = f32, size = 0.016 MB
|
| 37 |
+
[ 7/ 291] blk.0.attn_k.weight - [ 4096, 1024, 1, 1], type = f32, quantizing to q3_K .. size = 16.00 MiB -> 1.72 MiB | hist:
|
| 38 |
+
[ 8/ 291] blk.0.attn_output.weight - [ 4096, 4096, 1, 1], type = f32, quantizing to q4_K .. size = 64.00 MiB -> 9.00 MiB | hist:
|
| 39 |
+
[ 9/ 291] blk.0.attn_q.weight - [ 4096, 4096, 1, 1], type = f32, quantizing to q3_K .. size = 64.00 MiB -> 6.88 MiB | hist:
|
| 40 |
+
[ 10/ 291] blk.0.attn_v.weight - [ 4096, 1024, 1, 1], type = f32, quantizing to q5_K .. size = 16.00 MiB -> 2.75 MiB | hist:
|
| 41 |
+
[ 11/ 291] blk.1.attn_norm.weight - [ 4096, 1, 1, 1], type = f32, size = 0.016 MB
|
| 42 |
+
[ 12/ 291] blk.1.ffn_down.weight - [14336, 4096, 1, 1], type = f32, quantizing to q5_K .. size = 224.00 MiB -> 38.50 MiB | hist:
|
| 43 |
+
[ 13/ 291] blk.1.ffn_gate.weight - [ 4096, 14336, 1, 1], type = f32, quantizing to q3_K .. size = 224.00 MiB -> 24.06 MiB | hist:
|
| 44 |
+
[ 14/ 291] blk.1.ffn_up.weight - [ 4096, 14336, 1, 1], type = f32, quantizing to q3_K .. size = 224.00 MiB -> 24.06 MiB | hist:
|
| 45 |
+
[ 15/ 291] blk.1.ffn_norm.weight - [ 4096, 1, 1, 1], type = f32, size = 0.016 MB
|
| 46 |
+
[ 16/ 291] blk.1.attn_k.weight - [ 4096, 1024, 1, 1], type = f32, quantizing to q3_K .. size = 16.00 MiB -> 1.72 MiB | hist:
|
| 47 |
+
[ 17/ 291] blk.1.attn_output.weight - [ 4096, 4096, 1, 1], type = f32, quantizing to q4_K .. size = 64.00 MiB -> 9.00 MiB | hist:
|
| 48 |
+
[ 18/ 291] blk.1.attn_q.weight - [ 4096, 4096, 1, 1], type = f32, quantizing to q3_K .. size = 64.00 MiB -> 6.88 MiB | hist:
|
| 49 |
+
[ 19/ 291] blk.1.attn_v.weight - [ 4096, 1024, 1, 1], type = f32, quantizing to q5_K .. size = 16.00 MiB -> 2.75 MiB | hist:
|
| 50 |
+
[ 20/ 291] blk.2.attn_norm.weight - [ 4096, 1, 1, 1], type = f32, size = 0.016 MB
|
| 51 |
+
[ 21/ 291] blk.2.ffn_down.weight - [14336, 4096, 1, 1], type = f32, quantizing to q4_K .. size = 224.00 MiB -> 31.50 MiB | hist:
|
| 52 |
+
[ 22/ 291] blk.2.ffn_gate.weight - [ 4096, 14336, 1, 1], type = f32, quantizing to q3_K .. size = 224.00 MiB -> 24.06 MiB | hist:
|
| 53 |
+
[ 23/ 291] blk.2.ffn_up.weight - [ 4096, 14336, 1, 1], type = f32, quantizing to q3_K .. size = 224.00 MiB -> 24.06 MiB | hist:
|
| 54 |
+
[ 24/ 291] blk.2.ffn_norm.weight - [ 4096, 1, 1, 1], type = f32, size = 0.016 MB
|
| 55 |
+
[ 25/ 291] blk.2.attn_k.weight - [ 4096, 1024, 1, 1], type = f32, quantizing to q3_K .. size = 16.00 MiB -> 1.72 MiB | hist:
|
| 56 |
+
[ 26/ 291] blk.2.attn_output.weight - [ 4096, 4096, 1, 1], type = f32, quantizing to q4_K .. size = 64.00 MiB -> 9.00 MiB | hist:
|
| 57 |
+
[ 27/ 291] blk.2.attn_q.weight - [ 4096, 4096, 1, 1], type = f32, quantizing to q3_K .. size = 64.00 MiB -> 6.88 MiB | hist:
|
| 58 |
+
[ 28/ 291] blk.2.attn_v.weight - [ 4096, 1024, 1, 1], type = f32, quantizing to q4_K .. size = 16.00 MiB -> 2.25 MiB | hist:
|
| 59 |
+
[ 29/ 291] blk.3.ffn_gate.weight - [ 4096, 14336, 1, 1], type = f32, quantizing to q3_K .. size = 224.00 MiB -> 24.06 MiB | hist:
|
| 60 |
+
[ 30/ 291] blk.3.ffn_up.weight - [ 4096, 14336, 1, 1], type = f32, quantizing to q3_K .. size = 224.00 MiB -> 24.06 MiB | hist:
|
| 61 |
+
[ 31/ 291] blk.3.attn_k.weight - [ 4096, 1024, 1, 1], type = f32, quantizing to q3_K .. size = 16.00 MiB -> 1.72 MiB | hist:
|
| 62 |
+
[ 32/ 291] blk.3.attn_output.weight - [ 4096, 4096, 1, 1], type = f32, quantizing to q4_K .. size = 64.00 MiB -> 9.00 MiB | hist:
|
| 63 |
+
[ 33/ 291] blk.3.attn_q.weight - [ 4096, 4096, 1, 1], type = f32, quantizing to q3_K .. size = 64.00 MiB -> 6.88 MiB | hist:
|
| 64 |
+
[ 34/ 291] blk.3.attn_v.weight - [ 4096, 1024, 1, 1], type = f32, quantizing to q4_K .. size = 16.00 MiB -> 2.25 MiB | hist:
|
| 65 |
+
[ 35/ 291] blk.3.attn_norm.weight - [ 4096, 1, 1, 1], type = f32, size = 0.016 MB
|
| 66 |
+
[ 36/ 291] blk.3.ffn_down.weight - [14336, 4096, 1, 1], type = f32, quantizing to q4_K .. size = 224.00 MiB -> 31.50 MiB | hist:
|
| 67 |
+
[ 37/ 291] blk.3.ffn_norm.weight - [ 4096, 1, 1, 1], type = f32, size = 0.016 MB
|
| 68 |
+
[ 38/ 291] blk.4.attn_norm.weight - [ 4096, 1, 1, 1], type = f32, size = 0.016 MB
|
| 69 |
+
[ 39/ 291] blk.4.ffn_down.weight - [14336, 4096, 1, 1], type = f32, quantizing to q4_K .. size = 224.00 MiB -> 31.50 MiB | hist:
|
| 70 |
+
[ 40/ 291] blk.4.ffn_gate.weight - [ 4096, 14336, 1, 1], type = f32, quantizing to q3_K .. size = 224.00 MiB -> 24.06 MiB | hist:
|
| 71 |
+
[ 41/ 291] blk.4.ffn_up.weight - [ 4096, 14336, 1, 1], type = f32, quantizing to q3_K .. size = 224.00 MiB -> 24.06 MiB | hist:
|
| 72 |
+
[ 42/ 291] blk.4.ffn_norm.weight - [ 4096, 1, 1, 1], type = f32, size = 0.016 MB
|
| 73 |
+
[ 43/ 291] blk.4.attn_k.weight - [ 4096, 1024, 1, 1], type = f32, quantizing to q3_K .. size = 16.00 MiB -> 1.72 MiB | hist:
|
| 74 |
+
[ 44/ 291] blk.4.attn_output.weight - [ 4096, 4096, 1, 1], type = f32, quantizing to q4_K .. size = 64.00 MiB -> 9.00 MiB | hist:
|
| 75 |
+
[ 45/ 291] blk.4.attn_q.weight - [ 4096, 4096, 1, 1], type = f32, quantizing to q3_K .. size = 64.00 MiB -> 6.88 MiB | hist:
|
| 76 |
+
[ 46/ 291] blk.4.attn_v.weight - [ 4096, 1024, 1, 1], type = f32, quantizing to q4_K .. size = 16.00 MiB -> 2.25 MiB | hist:
|
| 77 |
+
[ 47/ 291] blk.5.attn_norm.weight - [ 4096, 1, 1, 1], type = f32, size = 0.016 MB
|
| 78 |
+
[ 48/ 291] blk.5.ffn_down.weight - [14336, 4096, 1, 1], type = f32, quantizing to q4_K .. size = 224.00 MiB -> 31.50 MiB | hist:
|
| 79 |
+
[ 49/ 291] blk.5.ffn_gate.weight - [ 4096, 14336, 1, 1], type = f32, quantizing to q3_K .. size = 224.00 MiB -> 24.06 MiB | hist:
|
| 80 |
+
[ 50/ 291] blk.5.ffn_up.weight - [ 4096, 14336, 1, 1], type = f32, quantizing to q3_K .. size = 224.00 MiB -> 24.06 MiB | hist:
|
| 81 |
+
[ 51/ 291] blk.5.ffn_norm.weight - [ 4096, 1, 1, 1], type = f32, size = 0.016 MB
|
| 82 |
+
[ 52/ 291] blk.5.attn_k.weight - [ 4096, 1024, 1, 1], type = f32, quantizing to q3_K .. size = 16.00 MiB -> 1.72 MiB | hist:
|
| 83 |
+
[ 53/ 291] blk.5.attn_output.weight - [ 4096, 4096, 1, 1], type = f32, quantizing to q4_K .. size = 64.00 MiB -> 9.00 MiB | hist:
|
| 84 |
+
[ 54/ 291] blk.5.attn_q.weight - [ 4096, 4096, 1, 1], type = f32, quantizing to q3_K .. size = 64.00 MiB -> 6.88 MiB | hist:
|
| 85 |
+
[ 55/ 291] blk.5.attn_v.weight - [ 4096, 1024, 1, 1], type = f32, quantizing to q4_K .. size = 16.00 MiB -> 2.25 MiB | hist:
|
| 86 |
+
[ 56/ 291] blk.6.attn_norm.weight - [ 4096, 1, 1, 1], type = f32, size = 0.016 MB
|
| 87 |
+
[ 57/ 291] blk.6.ffn_down.weight - [14336, 4096, 1, 1], type = f32, quantizing to q4_K .. size = 224.00 MiB -> 31.50 MiB | hist:
|
| 88 |
+
[ 58/ 291] blk.6.ffn_gate.weight - [ 4096, 14336, 1, 1], type = f32, quantizing to q3_K .. size = 224.00 MiB -> 24.06 MiB | hist:
|
| 89 |
+
[ 59/ 291] blk.6.ffn_up.weight - [ 4096, 14336, 1, 1], type = f32, quantizing to q3_K .. size = 224.00 MiB -> 24.06 MiB | hist:
|
| 90 |
+
[ 60/ 291] blk.6.ffn_norm.weight - [ 4096, 1, 1, 1], type = f32, size = 0.016 MB
|
| 91 |
+
[ 61/ 291] blk.6.attn_k.weight - [ 4096, 1024, 1, 1], type = f32, quantizing to q3_K .. size = 16.00 MiB -> 1.72 MiB | hist:
|
| 92 |
+
[ 62/ 291] blk.6.attn_output.weight - [ 4096, 4096, 1, 1], type = f32, quantizing to q4_K .. size = 64.00 MiB -> 9.00 MiB | hist:
|
| 93 |
+
[ 63/ 291] blk.6.attn_q.weight - [ 4096, 4096, 1, 1], type = f32, quantizing to q3_K .. size = 64.00 MiB -> 6.88 MiB | hist:
|
| 94 |
+
[ 64/ 291] blk.6.attn_v.weight - [ 4096, 1024, 1, 1], type = f32, quantizing to q4_K .. size = 16.00 MiB -> 2.25 MiB | hist:
|
| 95 |
+
[ 65/ 291] blk.7.attn_norm.weight - [ 4096, 1, 1, 1], type = f32, size = 0.016 MB
|
| 96 |
+
[ 66/ 291] blk.7.ffn_down.weight - [14336, 4096, 1, 1], type = f32, quantizing to q4_K .. size = 224.00 MiB -> 31.50 MiB | hist:
|
| 97 |
+
[ 67/ 291] blk.7.ffn_gate.weight - [ 4096, 14336, 1, 1], type = f32, quantizing to q3_K .. size = 224.00 MiB -> 24.06 MiB | hist:
|
| 98 |
+
[ 68/ 291] blk.7.ffn_up.weight - [ 4096, 14336, 1, 1], type = f32, quantizing to q3_K .. size = 224.00 MiB -> 24.06 MiB | hist:
|
| 99 |
+
[ 69/ 291] blk.7.ffn_norm.weight - [ 4096, 1, 1, 1], type = f32, size = 0.016 MB
|
| 100 |
+
[ 70/ 291] blk.7.attn_k.weight - [ 4096, 1024, 1, 1], type = f32, quantizing to q3_K .. size = 16.00 MiB -> 1.72 MiB | hist:
|
| 101 |
+
[ 71/ 291] blk.7.attn_output.weight - [ 4096, 4096, 1, 1], type = f32, quantizing to q4_K .. size = 64.00 MiB -> 9.00 MiB | hist:
|
| 102 |
+
[ 72/ 291] blk.7.attn_q.weight - [ 4096, 4096, 1, 1], type = f32, quantizing to q3_K .. size = 64.00 MiB -> 6.88 MiB | hist:
|
| 103 |
+
[ 73/ 291] blk.7.attn_v.weight - [ 4096, 1024, 1, 1], type = f32, quantizing to q4_K .. size = 16.00 MiB -> 2.25 MiB | hist:
|
| 104 |
+
[ 74/ 291] blk.8.attn_k.weight - [ 4096, 1024, 1, 1], type = f32, quantizing to q3_K .. size = 16.00 MiB -> 1.72 MiB | hist:
|
| 105 |
+
[ 75/ 291] blk.8.attn_output.weight - [ 4096, 4096, 1, 1], type = f32, quantizing to q4_K .. size = 64.00 MiB -> 9.00 MiB | hist:
|
| 106 |
+
[ 76/ 291] blk.8.attn_q.weight - [ 4096, 4096, 1, 1], type = f32, quantizing to q3_K .. size = 64.00 MiB -> 6.88 MiB | hist:
|
| 107 |
+
[ 77/ 291] blk.8.attn_v.weight - [ 4096, 1024, 1, 1], type = f32, quantizing to q4_K .. size = 16.00 MiB -> 2.25 MiB | hist:
|
| 108 |
+
[ 78/ 291] blk.10.attn_norm.weight - [ 4096, 1, 1, 1], type = f32, size = 0.016 MB
|
| 109 |
+
[ 79/ 291] blk.10.ffn_down.weight - [14336, 4096, 1, 1], type = f32, quantizing to q4_K .. size = 224.00 MiB -> 31.50 MiB | hist:
|
| 110 |
+
[ 80/ 291] blk.10.ffn_gate.weight - [ 4096, 14336, 1, 1], type = f32, quantizing to q3_K .. size = 224.00 MiB -> 24.06 MiB | hist:
|
| 111 |
+
[ 81/ 291] blk.10.ffn_up.weight - [ 4096, 14336, 1, 1], type = f32, quantizing to q3_K .. size = 224.00 MiB -> 24.06 MiB | hist:
|
| 112 |
+
[ 82/ 291] blk.10.ffn_norm.weight - [ 4096, 1, 1, 1], type = f32, size = 0.016 MB
|
| 113 |
+
[ 83/ 291] blk.10.attn_k.weight - [ 4096, 1024, 1, 1], type = f32, quantizing to q3_K .. size = 16.00 MiB -> 1.72 MiB | hist:
|
| 114 |
+
[ 84/ 291] blk.10.attn_output.weight - [ 4096, 4096, 1, 1], type = f32, quantizing to q4_K .. size = 64.00 MiB -> 9.00 MiB | hist:
|
| 115 |
+
[ 85/ 291] blk.10.attn_q.weight - [ 4096, 4096, 1, 1], type = f32, quantizing to q3_K .. size = 64.00 MiB -> 6.88 MiB | hist:
|
| 116 |
+
[ 86/ 291] blk.10.attn_v.weight - [ 4096, 1024, 1, 1], type = f32, quantizing to q4_K .. size = 16.00 MiB -> 2.25 MiB | hist:
|
| 117 |
+
[ 87/ 291] blk.11.attn_norm.weight - [ 4096, 1, 1, 1], type = f32, size = 0.016 MB
|
| 118 |
+
[ 88/ 291] blk.11.ffn_down.weight - [14336, 4096, 1, 1], type = f32, quantizing to q4_K .. size = 224.00 MiB -> 31.50 MiB | hist:
|
| 119 |
+
[ 89/ 291] blk.11.ffn_gate.weight - [ 4096, 14336, 1, 1], type = f32, quantizing to q3_K .. size = 224.00 MiB -> 24.06 MiB | hist:
|
| 120 |
+
[ 90/ 291] blk.11.ffn_up.weight - [ 4096, 14336, 1, 1], type = f32, quantizing to q3_K .. size = 224.00 MiB -> 24.06 MiB | hist:
|
| 121 |
+
[ 91/ 291] blk.11.ffn_norm.weight - [ 4096, 1, 1, 1], type = f32, size = 0.016 MB
|
| 122 |
+
[ 92/ 291] blk.11.attn_k.weight - [ 4096, 1024, 1, 1], type = f32, quantizing to q3_K .. size = 16.00 MiB -> 1.72 MiB | hist:
|
| 123 |
+
[ 93/ 291] blk.11.attn_output.weight - [ 4096, 4096, 1, 1], type = f32, quantizing to q4_K .. size = 64.00 MiB -> 9.00 MiB | hist:
|
| 124 |
+
[ 94/ 291] blk.11.attn_q.weight - [ 4096, 4096, 1, 1], type = f32, quantizing to q3_K .. size = 64.00 MiB -> 6.88 MiB | hist:
|
| 125 |
+
[ 95/ 291] blk.11.attn_v.weight - [ 4096, 1024, 1, 1], type = f32, quantizing to q4_K .. size = 16.00 MiB -> 2.25 MiB | hist:
|
| 126 |
+
[ 96/ 291] blk.12.ffn_gate.weight - [ 4096, 14336, 1, 1], type = f32, quantizing to q3_K .. size = 224.00 MiB -> 24.06 MiB | hist:
|
| 127 |
+
[ 97/ 291] blk.12.ffn_up.weight - [ 4096, 14336, 1, 1], type = f32, quantizing to q3_K .. size = 224.00 MiB -> 24.06 MiB | hist:
|
| 128 |
+
[ 98/ 291] blk.12.attn_k.weight - [ 4096, 1024, 1, 1], type = f32, quantizing to q3_K .. size = 16.00 MiB -> 1.72 MiB | hist:
|
| 129 |
+
[ 99/ 291] blk.12.attn_output.weight - [ 4096, 4096, 1, 1], type = f32, quantizing to q4_K .. size = 64.00 MiB -> 9.00 MiB | hist:
|
| 130 |
+
[ 100/ 291] blk.12.attn_q.weight - [ 4096, 4096, 1, 1], type = f32, quantizing to q3_K .. size = 64.00 MiB -> 6.88 MiB | hist:
|
| 131 |
+
[ 101/ 291] blk.12.attn_v.weight - [ 4096, 1024, 1, 1], type = f32, quantizing to q4_K .. size = 16.00 MiB -> 2.25 MiB | hist:
|
| 132 |
+
[ 102/ 291] blk.8.attn_norm.weight - [ 4096, 1, 1, 1], type = f32, size = 0.016 MB
|
| 133 |
+
[ 103/ 291] blk.8.ffn_down.weight - [14336, 4096, 1, 1], type = f32, quantizing to q4_K .. size = 224.00 MiB -> 31.50 MiB | hist:
|
| 134 |
+
[ 104/ 291] blk.8.ffn_gate.weight - [ 4096, 14336, 1, 1], type = f32, quantizing to q3_K .. size = 224.00 MiB -> 24.06 MiB | hist:
|
| 135 |
+
[ 105/ 291] blk.8.ffn_up.weight - [ 4096, 14336, 1, 1], type = f32, quantizing to q3_K .. size = 224.00 MiB -> 24.06 MiB | hist:
|
| 136 |
+
[ 106/ 291] blk.8.ffn_norm.weight - [ 4096, 1, 1, 1], type = f32, size = 0.016 MB
|
| 137 |
+
[ 107/ 291] blk.9.attn_norm.weight - [ 4096, 1, 1, 1], type = f32, size = 0.016 MB
|
| 138 |
+
[ 108/ 291] blk.9.ffn_down.weight - [14336, 4096, 1, 1], type = f32, quantizing to q4_K .. size = 224.00 MiB -> 31.50 MiB | hist:
|
| 139 |
+
[ 109/ 291] blk.9.ffn_gate.weight - [ 4096, 14336, 1, 1], type = f32, quantizing to q3_K .. size = 224.00 MiB -> 24.06 MiB | hist:
|
| 140 |
+
[ 110/ 291] blk.9.ffn_up.weight - [ 4096, 14336, 1, 1], type = f32, quantizing to q3_K .. size = 224.00 MiB -> 24.06 MiB | hist:
|
| 141 |
+
[ 111/ 291] blk.9.ffn_norm.weight - [ 4096, 1, 1, 1], type = f32, size = 0.016 MB
|
| 142 |
+
[ 112/ 291] blk.9.attn_k.weight - [ 4096, 1024, 1, 1], type = f32, quantizing to q3_K .. size = 16.00 MiB -> 1.72 MiB | hist:
|
| 143 |
+
[ 113/ 291] blk.9.attn_output.weight - [ 4096, 4096, 1, 1], type = f32, quantizing to q4_K .. size = 64.00 MiB -> 9.00 MiB | hist:
|
| 144 |
+
[ 114/ 291] blk.9.attn_q.weight - [ 4096, 4096, 1, 1], type = f32, quantizing to q3_K .. size = 64.00 MiB -> 6.88 MiB | hist:
|
| 145 |
+
[ 115/ 291] blk.9.attn_v.weight - [ 4096, 1024, 1, 1], type = f32, quantizing to q4_K .. size = 16.00 MiB -> 2.25 MiB | hist:
|
| 146 |
+
[ 116/ 291] blk.12.attn_norm.weight - [ 4096, 1, 1, 1], type = f32, size = 0.016 MB
|
| 147 |
+
[ 117/ 291] blk.12.ffn_down.weight - [14336, 4096, 1, 1], type = f32, quantizing to q4_K .. size = 224.00 MiB -> 31.50 MiB | hist:
|
| 148 |
+
[ 118/ 291] blk.12.ffn_norm.weight - [ 4096, 1, 1, 1], type = f32, size = 0.016 MB
|
| 149 |
+
[ 119/ 291] blk.13.attn_norm.weight - [ 4096, 1, 1, 1], type = f32, size = 0.016 MB
|
| 150 |
+
[ 120/ 291] blk.13.ffn_down.weight - [14336, 4096, 1, 1], type = f32, quantizing to q4_K .. size = 224.00 MiB -> 31.50 MiB | hist:
|
| 151 |
+
[ 121/ 291] blk.13.ffn_gate.weight - [ 4096, 14336, 1, 1], type = f32, quantizing to q3_K .. size = 224.00 MiB -> 24.06 MiB | hist:
|
| 152 |
+
[ 122/ 291] blk.13.ffn_up.weight - [ 4096, 14336, 1, 1], type = f32, quantizing to q3_K .. size = 224.00 MiB -> 24.06 MiB | hist:
|
| 153 |
+
[ 123/ 291] blk.13.ffn_norm.weight - [ 4096, 1, 1, 1], type = f32, size = 0.016 MB
|
| 154 |
+
[ 124/ 291] blk.13.attn_k.weight - [ 4096, 1024, 1, 1], type = f32, quantizing to q3_K .. size = 16.00 MiB -> 1.72 MiB | hist:
|
| 155 |
+
[ 125/ 291] blk.13.attn_output.weight - [ 4096, 4096, 1, 1], type = f32, quantizing to q4_K .. size = 64.00 MiB -> 9.00 MiB | hist:
|
| 156 |
+
[ 126/ 291] blk.13.attn_q.weight - [ 4096, 4096, 1, 1], type = f32, quantizing to q3_K .. size = 64.00 MiB -> 6.88 MiB | hist:
|
| 157 |
+
[ 127/ 291] blk.13.attn_v.weight - [ 4096, 1024, 1, 1], type = f32, quantizing to q4_K .. size = 16.00 MiB -> 2.25 MiB | hist:
|
| 158 |
+
[ 128/ 291] blk.14.attn_norm.weight - [ 4096, 1, 1, 1], type = f32, size = 0.016 MB
|
| 159 |
+
[ 129/ 291] blk.14.ffn_down.weight - [14336, 4096, 1, 1], type = f32, quantizing to q4_K .. size = 224.00 MiB -> 31.50 MiB | hist:
|
| 160 |
+
[ 130/ 291] blk.14.ffn_gate.weight - [ 4096, 14336, 1, 1], type = f32, quantizing to q3_K .. size = 224.00 MiB -> 24.06 MiB | hist:
|
| 161 |
+
[ 131/ 291] blk.14.ffn_up.weight - [ 4096, 14336, 1, 1], type = f32, quantizing to q3_K .. size = 224.00 MiB -> 24.06 MiB | hist:
|
| 162 |
+
[ 132/ 291] blk.14.ffn_norm.weight - [ 4096, 1, 1, 1], type = f32, size = 0.016 MB
|
| 163 |
+
[ 133/ 291] blk.14.attn_k.weight - [ 4096, 1024, 1, 1], type = f32, quantizing to q3_K .. size = 16.00 MiB -> 1.72 MiB | hist:
|
| 164 |
+
[ 134/ 291] blk.14.attn_output.weight - [ 4096, 4096, 1, 1], type = f32, quantizing to q4_K .. size = 64.00 MiB -> 9.00 MiB | hist:
|
| 165 |
+
[ 135/ 291] blk.14.attn_q.weight - [ 4096, 4096, 1, 1], type = f32, quantizing to q3_K .. size = 64.00 MiB -> 6.88 MiB | hist:
|
| 166 |
+
[ 136/ 291] blk.14.attn_v.weight - [ 4096, 1024, 1, 1], type = f32, quantizing to q4_K .. size = 16.00 MiB -> 2.25 MiB | hist:
|
| 167 |
+
[ 137/ 291] blk.15.attn_norm.weight - [ 4096, 1, 1, 1], type = f32, size = 0.016 MB
|
| 168 |
+
[ 138/ 291] blk.15.ffn_down.weight - [14336, 4096, 1, 1], type = f32, quantizing to q4_K .. size = 224.00 MiB -> 31.50 MiB | hist:
|
| 169 |
+
[ 139/ 291] blk.15.ffn_gate.weight - [ 4096, 14336, 1, 1], type = f32, quantizing to q3_K .. size = 224.00 MiB -> 24.06 MiB | hist:
|
| 170 |
+
[ 140/ 291] blk.15.ffn_up.weight - [ 4096, 14336, 1, 1], type = f32, quantizing to q3_K .. size = 224.00 MiB -> 24.06 MiB | hist:
|
| 171 |
+
[ 141/ 291] blk.15.ffn_norm.weight - [ 4096, 1, 1, 1], type = f32, size = 0.016 MB
|
| 172 |
+
[ 142/ 291] blk.15.attn_k.weight - [ 4096, 1024, 1, 1], type = f32, quantizing to q3_K .. size = 16.00 MiB -> 1.72 MiB | hist:
|
| 173 |
+
[ 143/ 291] blk.15.attn_output.weight - [ 4096, 4096, 1, 1], type = f32, quantizing to q4_K .. size = 64.00 MiB -> 9.00 MiB | hist:
|
| 174 |
+
[ 144/ 291] blk.15.attn_q.weight - [ 4096, 4096, 1, 1], type = f32, quantizing to q3_K .. size = 64.00 MiB -> 6.88 MiB | hist:
|
| 175 |
+
[ 145/ 291] blk.15.attn_v.weight - [ 4096, 1024, 1, 1], type = f32, quantizing to q4_K .. size = 16.00 MiB -> 2.25 MiB | hist:
|
| 176 |
+
[ 146/ 291] blk.16.attn_norm.weight - [ 4096, 1, 1, 1], type = f32, size = 0.016 MB
|
| 177 |
+
[ 147/ 291] blk.16.ffn_down.weight - [14336, 4096, 1, 1], type = f32, quantizing to q4_K .. size = 224.00 MiB -> 31.50 MiB | hist:
|
| 178 |
+
[ 148/ 291] blk.16.ffn_gate.weight - [ 4096, 14336, 1, 1], type = f32, quantizing to q3_K .. size = 224.00 MiB -> 24.06 MiB | hist:
|
| 179 |
+
[ 149/ 291] blk.16.ffn_up.weight - [ 4096, 14336, 1, 1], type = f32, quantizing to q3_K .. size = 224.00 MiB -> 24.06 MiB | hist:
|
| 180 |
+
[ 150/ 291] blk.16.ffn_norm.weight - [ 4096, 1, 1, 1], type = f32, size = 0.016 MB
|
| 181 |
+
[ 151/ 291] blk.16.attn_k.weight - [ 4096, 1024, 1, 1], type = f32, quantizing to q3_K .. size = 16.00 MiB -> 1.72 MiB | hist:
|
| 182 |
+
[ 152/ 291] blk.16.attn_output.weight - [ 4096, 4096, 1, 1], type = f32, quantizing to q4_K .. size = 64.00 MiB -> 9.00 MiB | hist:
|
| 183 |
+
[ 153/ 291] blk.16.attn_q.weight - [ 4096, 4096, 1, 1], type = f32, quantizing to q3_K .. size = 64.00 MiB -> 6.88 MiB | hist:
|
| 184 |
+
[ 154/ 291] blk.16.attn_v.weight - [ 4096, 1024, 1, 1], type = f32, quantizing to q4_K .. size = 16.00 MiB -> 2.25 MiB | hist:
|
| 185 |
+
[ 155/ 291] blk.17.attn_k.weight - [ 4096, 1024, 1, 1], type = f32, quantizing to q3_K .. size = 16.00 MiB -> 1.72 MiB | hist:
|
| 186 |
+
[ 156/ 291] blk.17.attn_output.weight - [ 4096, 4096, 1, 1], type = f32, quantizing to q4_K .. size = 64.00 MiB -> 9.00 MiB | hist:
|
| 187 |
+
[ 157/ 291] blk.17.attn_q.weight - [ 4096, 4096, 1, 1], type = f32, quantizing to q3_K .. size = 64.00 MiB -> 6.88 MiB | hist:
|
| 188 |
+
[ 158/ 291] blk.17.attn_v.weight - [ 4096, 1024, 1, 1], type = f32, quantizing to q4_K .. size = 16.00 MiB -> 2.25 MiB | hist:
|
| 189 |
+
[ 159/ 291] blk.17.attn_norm.weight - [ 4096, 1, 1, 1], type = f32, size = 0.016 MB
|
| 190 |
+
[ 160/ 291] blk.17.ffn_down.weight - [14336, 4096, 1, 1], type = f32, quantizing to q4_K .. size = 224.00 MiB -> 31.50 MiB | hist:
|
| 191 |
+
[ 161/ 291] blk.17.ffn_gate.weight - [ 4096, 14336, 1, 1], type = f32, quantizing to q3_K .. size = 224.00 MiB -> 24.06 MiB | hist:
|
| 192 |
+
[ 162/ 291] blk.17.ffn_up.weight - [ 4096, 14336, 1, 1], type = f32, quantizing to q3_K .. size = 224.00 MiB -> 24.06 MiB | hist:
|
| 193 |
+
[ 163/ 291] blk.17.ffn_norm.weight - [ 4096, 1, 1, 1], type = f32, size = 0.016 MB
|
| 194 |
+
[ 164/ 291] blk.18.attn_norm.weight - [ 4096, 1, 1, 1], type = f32, size = 0.016 MB
|
| 195 |
+
[ 165/ 291] blk.18.ffn_down.weight - [14336, 4096, 1, 1], type = f32, quantizing to q4_K .. size = 224.00 MiB -> 31.50 MiB | hist:
|
| 196 |
+
[ 166/ 291] blk.18.ffn_gate.weight - [ 4096, 14336, 1, 1], type = f32, quantizing to q3_K .. size = 224.00 MiB -> 24.06 MiB | hist:
|
| 197 |
+
[ 167/ 291] blk.18.ffn_up.weight - [ 4096, 14336, 1, 1], type = f32, quantizing to q3_K .. size = 224.00 MiB -> 24.06 MiB | hist:
|
| 198 |
+
[ 168/ 291] blk.18.ffn_norm.weight - [ 4096, 1, 1, 1], type = f32, size = 0.016 MB
|
| 199 |
+
[ 169/ 291] blk.18.attn_k.weight - [ 4096, 1024, 1, 1], type = f32, quantizing to q3_K .. size = 16.00 MiB -> 1.72 MiB | hist:
|
| 200 |
+
[ 170/ 291] blk.18.attn_output.weight - [ 4096, 4096, 1, 1], type = f32, quantizing to q4_K .. size = 64.00 MiB -> 9.00 MiB | hist:
|
| 201 |
+
[ 171/ 291] blk.18.attn_q.weight - [ 4096, 4096, 1, 1], type = f32, quantizing to q3_K .. size = 64.00 MiB -> 6.88 MiB | hist:
|
| 202 |
+
[ 172/ 291] blk.18.attn_v.weight - [ 4096, 1024, 1, 1], type = f32, quantizing to q4_K .. size = 16.00 MiB -> 2.25 MiB | hist:
|
| 203 |
+
[ 173/ 291] blk.19.attn_norm.weight - [ 4096, 1, 1, 1], type = f32, size = 0.016 MB
|
| 204 |
+
[ 174/ 291] blk.19.ffn_down.weight - [14336, 4096, 1, 1], type = f32, quantizing to q4_K .. size = 224.00 MiB -> 31.50 MiB | hist:
|
| 205 |
+
[ 175/ 291] blk.19.ffn_gate.weight - [ 4096, 14336, 1, 1], type = f32, quantizing to q3_K .. size = 224.00 MiB -> 24.06 MiB | hist:
|
| 206 |
+
[ 176/ 291] blk.19.ffn_up.weight - [ 4096, 14336, 1, 1], type = f32, quantizing to q3_K .. size = 224.00 MiB -> 24.06 MiB | hist:
|
| 207 |
+
[ 177/ 291] blk.19.ffn_norm.weight - [ 4096, 1, 1, 1], type = f32, size = 0.016 MB
|
| 208 |
+
[ 178/ 291] blk.19.attn_k.weight - [ 4096, 1024, 1, 1], type = f32, quantizing to q3_K .. size = 16.00 MiB -> 1.72 MiB | hist:
|
| 209 |
+
[ 179/ 291] blk.19.attn_output.weight - [ 4096, 4096, 1, 1], type = f32, quantizing to q4_K .. size = 64.00 MiB -> 9.00 MiB | hist:
|
| 210 |
+
[ 180/ 291] blk.19.attn_q.weight - [ 4096, 4096, 1, 1], type = f32, quantizing to q3_K .. size = 64.00 MiB -> 6.88 MiB | hist:
|
| 211 |
+
[ 181/ 291] blk.19.attn_v.weight - [ 4096, 1024, 1, 1], type = f32, quantizing to q4_K .. size = 16.00 MiB -> 2.25 MiB | hist:
|
| 212 |
+
[ 182/ 291] blk.20.attn_norm.weight - [ 4096, 1, 1, 1], type = f32, size = 0.016 MB
|
| 213 |
+
[ 183/ 291] blk.20.ffn_down.weight - [14336, 4096, 1, 1], type = f32, quantizing to q4_K .. size = 224.00 MiB -> 31.50 MiB | hist:
|
| 214 |
+
[ 184/ 291] blk.20.ffn_gate.weight - [ 4096, 14336, 1, 1], type = f32, quantizing to q3_K .. size = 224.00 MiB -> 24.06 MiB | hist:
|
| 215 |
+
[ 185/ 291] blk.20.ffn_up.weight - [ 4096, 14336, 1, 1], type = f32, quantizing to q3_K .. size = 224.00 MiB -> 24.06 MiB | hist:
|
| 216 |
+
[ 186/ 291] blk.20.ffn_norm.weight - [ 4096, 1, 1, 1], type = f32, size = 0.016 MB
|
| 217 |
+
[ 187/ 291] blk.20.attn_k.weight - [ 4096, 1024, 1, 1], type = f32, quantizing to q3_K .. size = 16.00 MiB -> 1.72 MiB | hist:
|
| 218 |
+
[ 188/ 291] blk.20.attn_output.weight - [ 4096, 4096, 1, 1], type = f32, quantizing to q4_K .. size = 64.00 MiB -> 9.00 MiB | hist:
|
| 219 |
+
[ 189/ 291] blk.20.attn_q.weight - [ 4096, 4096, 1, 1], type = f32, quantizing to q3_K .. size = 64.00 MiB -> 6.88 MiB | hist:
|
| 220 |
+
[ 190/ 291] blk.20.attn_v.weight - [ 4096, 1024, 1, 1], type = f32, quantizing to q4_K .. size = 16.00 MiB -> 2.25 MiB | hist:
|
| 221 |
+
[ 191/ 291] blk.21.ffn_gate.weight - [ 4096, 14336, 1, 1], type = f32, quantizing to q3_K .. size = 224.00 MiB -> 24.06 MiB | hist:
|
| 222 |
+
[ 192/ 291] blk.21.ffn_up.weight - [ 4096, 14336, 1, 1], type = f32, quantizing to q3_K .. size = 224.00 MiB -> 24.06 MiB | hist:
|
| 223 |
+
[ 193/ 291] blk.21.attn_k.weight - [ 4096, 1024, 1, 1], type = f32, quantizing to q3_K .. size = 16.00 MiB -> 1.72 MiB | hist:
|
| 224 |
+
[ 194/ 291] blk.21.attn_output.weight - [ 4096, 4096, 1, 1], type = f32, quantizing to q4_K .. size = 64.00 MiB -> 9.00 MiB | hist:
|
| 225 |
+
[ 195/ 291] blk.21.attn_q.weight - [ 4096, 4096, 1, 1], type = f32, quantizing to q3_K .. size = 64.00 MiB -> 6.88 MiB | hist:
|
| 226 |
+
[ 196/ 291] blk.21.attn_v.weight - [ 4096, 1024, 1, 1], type = f32, quantizing to q4_K .. size = 16.00 MiB -> 2.25 MiB | hist:
|
| 227 |
+
[ 197/ 291] blk.21.attn_norm.weight - [ 4096, 1, 1, 1], type = f32, size = 0.016 MB
|
| 228 |
+
[ 198/ 291] blk.21.ffn_down.weight - [14336, 4096, 1, 1], type = f32, quantizing to q4_K .. size = 224.00 MiB -> 31.50 MiB | hist:
|
| 229 |
+
[ 199/ 291] blk.21.ffn_norm.weight - [ 4096, 1, 1, 1], type = f32, size = 0.016 MB
|
| 230 |
+
[ 200/ 291] blk.22.attn_norm.weight - [ 4096, 1, 1, 1], type = f32, size = 0.016 MB
|
| 231 |
+
[ 201/ 291] blk.22.ffn_down.weight - [14336, 4096, 1, 1], type = f32, quantizing to q4_K .. size = 224.00 MiB -> 31.50 MiB | hist:
|
| 232 |
+
[ 202/ 291] blk.22.ffn_gate.weight - [ 4096, 14336, 1, 1], type = f32, quantizing to q3_K .. size = 224.00 MiB -> 24.06 MiB | hist:
|
| 233 |
+
[ 203/ 291] blk.22.ffn_up.weight - [ 4096, 14336, 1, 1], type = f32, quantizing to q3_K .. size = 224.00 MiB -> 24.06 MiB | hist:
|
| 234 |
+
[ 204/ 291] blk.22.ffn_norm.weight - [ 4096, 1, 1, 1], type = f32, size = 0.016 MB
|
| 235 |
+
[ 205/ 291] blk.22.attn_k.weight - [ 4096, 1024, 1, 1], type = f32, quantizing to q3_K .. size = 16.00 MiB -> 1.72 MiB | hist:
|
| 236 |
+
[ 206/ 291] blk.22.attn_output.weight - [ 4096, 4096, 1, 1], type = f32, quantizing to q4_K .. size = 64.00 MiB -> 9.00 MiB | hist:
|
| 237 |
+
[ 207/ 291] blk.22.attn_q.weight - [ 4096, 4096, 1, 1], type = f32, quantizing to q3_K .. size = 64.00 MiB -> 6.88 MiB | hist:
|
| 238 |
+
[ 208/ 291] blk.22.attn_v.weight - [ 4096, 1024, 1, 1], type = f32, quantizing to q4_K .. size = 16.00 MiB -> 2.25 MiB | hist:
|
| 239 |
+
[ 209/ 291] blk.23.attn_norm.weight - [ 4096, 1, 1, 1], type = f32, size = 0.016 MB
|
| 240 |
+
[ 210/ 291] blk.23.ffn_down.weight - [14336, 4096, 1, 1], type = f32, quantizing to q4_K .. size = 224.00 MiB -> 31.50 MiB | hist:
|
| 241 |
+
[ 211/ 291] blk.23.ffn_gate.weight - [ 4096, 14336, 1, 1], type = f32, quantizing to q3_K .. size = 224.00 MiB -> 24.06 MiB | hist:
|
| 242 |
+
[ 212/ 291] blk.23.ffn_up.weight - [ 4096, 14336, 1, 1], type = f32, quantizing to q3_K .. size = 224.00 MiB -> 24.06 MiB | hist:
|
| 243 |
+
[ 213/ 291] blk.23.ffn_norm.weight - [ 4096, 1, 1, 1], type = f32, size = 0.016 MB
|
| 244 |
+
[ 214/ 291] blk.23.attn_k.weight - [ 4096, 1024, 1, 1], type = f32, quantizing to q3_K .. size = 16.00 MiB -> 1.72 MiB | hist:
|
| 245 |
+
[ 215/ 291] blk.23.attn_output.weight - [ 4096, 4096, 1, 1], type = f32, quantizing to q4_K .. size = 64.00 MiB -> 9.00 MiB | hist:
|
| 246 |
+
[ 216/ 291] blk.23.attn_q.weight - [ 4096, 4096, 1, 1], type = f32, quantizing to q3_K .. size = 64.00 MiB -> 6.88 MiB | hist:
|
| 247 |
+
[ 217/ 291] blk.23.attn_v.weight - [ 4096, 1024, 1, 1], type = f32, quantizing to q4_K .. size = 16.00 MiB -> 2.25 MiB | hist:
|
| 248 |
+
[ 218/ 291] blk.24.attn_norm.weight - [ 4096, 1, 1, 1], type = f32, size = 0.016 MB
|
| 249 |
+
[ 219/ 291] blk.24.ffn_down.weight - [14336, 4096, 1, 1], type = f32, quantizing to q4_K .. size = 224.00 MiB -> 31.50 MiB | hist:
|
| 250 |
+
[ 220/ 291] blk.24.ffn_gate.weight - [ 4096, 14336, 1, 1], type = f32, quantizing to q3_K .. size = 224.00 MiB -> 24.06 MiB | hist:
|
| 251 |
+
[ 221/ 291] blk.24.ffn_up.weight - [ 4096, 14336, 1, 1], type = f32, quantizing to q3_K .. size = 224.00 MiB -> 24.06 MiB | hist:
|
| 252 |
+
[ 222/ 291] blk.24.ffn_norm.weight - [ 4096, 1, 1, 1], type = f32, size = 0.016 MB
|
| 253 |
+
[ 223/ 291] blk.24.attn_k.weight - [ 4096, 1024, 1, 1], type = f32, quantizing to q3_K .. size = 16.00 MiB -> 1.72 MiB | hist:
|
| 254 |
+
[ 224/ 291] blk.24.attn_output.weight - [ 4096, 4096, 1, 1], type = f32, quantizing to q4_K .. size = 64.00 MiB -> 9.00 MiB | hist:
|
| 255 |
+
[ 225/ 291] blk.24.attn_q.weight - [ 4096, 4096, 1, 1], type = f32, quantizing to q3_K .. size = 64.00 MiB -> 6.88 MiB | hist:
|
| 256 |
+
[ 226/ 291] blk.24.attn_v.weight - [ 4096, 1024, 1, 1], type = f32, quantizing to q4_K .. size = 16.00 MiB -> 2.25 MiB | hist:
|
| 257 |
+
[ 227/ 291] blk.25.attn_norm.weight - [ 4096, 1, 1, 1], type = f32, size = 0.016 MB
|
| 258 |
+
[ 228/ 291] blk.25.ffn_down.weight - [14336, 4096, 1, 1], type = f32, quantizing to q4_K .. size = 224.00 MiB -> 31.50 MiB | hist:
|
| 259 |
+
[ 229/ 291] blk.25.ffn_gate.weight - [ 4096, 14336, 1, 1], type = f32, quantizing to q3_K .. size = 224.00 MiB -> 24.06 MiB | hist:
|
| 260 |
+
[ 230/ 291] blk.25.ffn_up.weight - [ 4096, 14336, 1, 1], type = f32, quantizing to q3_K .. size = 224.00 MiB -> 24.06 MiB | hist:
|
| 261 |
+
[ 231/ 291] blk.25.ffn_norm.weight - [ 4096, 1, 1, 1], type = f32, size = 0.016 MB
|
| 262 |
+
[ 232/ 291] blk.25.attn_k.weight - [ 4096, 1024, 1, 1], type = f32, quantizing to q3_K .. size = 16.00 MiB -> 1.72 MiB | hist:
|
| 263 |
+
[ 233/ 291] blk.25.attn_output.weight - [ 4096, 4096, 1, 1], type = f32, quantizing to q4_K .. size = 64.00 MiB -> 9.00 MiB | hist:
|
| 264 |
+
[ 234/ 291] blk.25.attn_q.weight - [ 4096, 4096, 1, 1], type = f32, quantizing to q3_K .. size = 64.00 MiB -> 6.88 MiB | hist:
|
| 265 |
+
[ 235/ 291] blk.25.attn_v.weight - [ 4096, 1024, 1, 1], type = f32, quantizing to q4_K .. size = 16.00 MiB -> 2.25 MiB | hist:
|
| 266 |
+
[ 236/ 291] blk.26.attn_k.weight - [ 4096, 1024, 1, 1], type = f32, quantizing to q3_K .. size = 16.00 MiB -> 1.72 MiB | hist:
|
| 267 |
+
[ 237/ 291] blk.26.attn_output.weight - [ 4096, 4096, 1, 1], type = f32, quantizing to q4_K .. size = 64.00 MiB -> 9.00 MiB | hist:
|
| 268 |
+
[ 238/ 291] blk.26.attn_q.weight - [ 4096, 4096, 1, 1], type = f32, quantizing to q3_K .. size = 64.00 MiB -> 6.88 MiB | hist:
|
| 269 |
+
[ 239/ 291] blk.26.attn_v.weight - [ 4096, 1024, 1, 1], type = f32, quantizing to q4_K .. size = 16.00 MiB -> 2.25 MiB | hist:
|
| 270 |
+
[ 240/ 291] blk.26.attn_norm.weight - [ 4096, 1, 1, 1], type = f32, size = 0.016 MB
|
| 271 |
+
[ 241/ 291] blk.26.ffn_down.weight - [14336, 4096, 1, 1], type = f32, quantizing to q4_K .. size = 224.00 MiB -> 31.50 MiB | hist:
|
| 272 |
+
[ 242/ 291] blk.26.ffn_gate.weight - [ 4096, 14336, 1, 1], type = f32, quantizing to q3_K .. size = 224.00 MiB -> 24.06 MiB | hist:
|
| 273 |
+
[ 243/ 291] blk.26.ffn_up.weight - [ 4096, 14336, 1, 1], type = f32, quantizing to q3_K .. size = 224.00 MiB -> 24.06 MiB | hist:
|
| 274 |
+
[ 244/ 291] blk.26.ffn_norm.weight - [ 4096, 1, 1, 1], type = f32, size = 0.016 MB
|
| 275 |
+
[ 245/ 291] blk.27.attn_norm.weight - [ 4096, 1, 1, 1], type = f32, size = 0.016 MB
|
| 276 |
+
[ 246/ 291] blk.27.ffn_down.weight - [14336, 4096, 1, 1], type = f32, quantizing to q4_K .. size = 224.00 MiB -> 31.50 MiB | hist:
|
| 277 |
+
[ 247/ 291] blk.27.ffn_gate.weight - [ 4096, 14336, 1, 1], type = f32, quantizing to q3_K .. size = 224.00 MiB -> 24.06 MiB | hist:
|
| 278 |
+
[ 248/ 291] blk.27.ffn_up.weight - [ 4096, 14336, 1, 1], type = f32, quantizing to q3_K .. size = 224.00 MiB -> 24.06 MiB | hist:
|
| 279 |
+
[ 249/ 291] blk.27.ffn_norm.weight - [ 4096, 1, 1, 1], type = f32, size = 0.016 MB
|
| 280 |
+
[ 250/ 291] blk.27.attn_k.weight - [ 4096, 1024, 1, 1], type = f32, quantizing to q3_K .. size = 16.00 MiB -> 1.72 MiB | hist:
|
| 281 |
+
[ 251/ 291] blk.27.attn_output.weight - [ 4096, 4096, 1, 1], type = f32, quantizing to q4_K .. size = 64.00 MiB -> 9.00 MiB | hist:
|
| 282 |
+
[ 252/ 291] blk.27.attn_q.weight - [ 4096, 4096, 1, 1], type = f32, quantizing to q3_K .. size = 64.00 MiB -> 6.88 MiB | hist:
|
| 283 |
+
[ 253/ 291] blk.27.attn_v.weight - [ 4096, 1024, 1, 1], type = f32, quantizing to q4_K .. size = 16.00 MiB -> 2.25 MiB | hist:
|
| 284 |
+
[ 254/ 291] blk.28.attn_norm.weight - [ 4096, 1, 1, 1], type = f32, size = 0.016 MB
|
| 285 |
+
[ 255/ 291] blk.28.ffn_down.weight - [14336, 4096, 1, 1], type = f32, quantizing to q4_K .. size = 224.00 MiB -> 31.50 MiB | hist:
|
| 286 |
+
[ 256/ 291] blk.28.ffn_gate.weight - [ 4096, 14336, 1, 1], type = f32, quantizing to q3_K .. size = 224.00 MiB -> 24.06 MiB | hist:
|
| 287 |
+
[ 257/ 291] blk.28.ffn_up.weight - [ 4096, 14336, 1, 1], type = f32, quantizing to q3_K .. size = 224.00 MiB -> 24.06 MiB | hist:
|
| 288 |
+
[ 258/ 291] blk.28.ffn_norm.weight - [ 4096, 1, 1, 1], type = f32, size = 0.016 MB
|
| 289 |
+
[ 259/ 291] blk.28.attn_k.weight - [ 4096, 1024, 1, 1], type = f32, quantizing to q3_K .. size = 16.00 MiB -> 1.72 MiB | hist:
|
| 290 |
+
[ 260/ 291] blk.28.attn_output.weight - [ 4096, 4096, 1, 1], type = f32, quantizing to q4_K .. size = 64.00 MiB -> 9.00 MiB | hist:
|
| 291 |
+
[ 261/ 291] blk.28.attn_q.weight - [ 4096, 4096, 1, 1], type = f32, quantizing to q3_K .. size = 64.00 MiB -> 6.88 MiB | hist:
|
| 292 |
+
[ 262/ 291] blk.28.attn_v.weight - [ 4096, 1024, 1, 1], type = f32, quantizing to q4_K .. size = 16.00 MiB -> 2.25 MiB | hist:
|
| 293 |
+
[ 263/ 291] blk.29.attn_norm.weight - [ 4096, 1, 1, 1], type = f32, size = 0.016 MB
|
| 294 |
+
[ 264/ 291] blk.29.ffn_down.weight - [14336, 4096, 1, 1], type = f32, quantizing to q4_K .. size = 224.00 MiB -> 31.50 MiB | hist:
|
| 295 |
+
[ 265/ 291] blk.29.ffn_gate.weight - [ 4096, 14336, 1, 1], type = f32, quantizing to q3_K .. size = 224.00 MiB -> 24.06 MiB | hist:
|
| 296 |
+
[ 266/ 291] blk.29.ffn_up.weight - [ 4096, 14336, 1, 1], type = f32, quantizing to q3_K .. size = 224.00 MiB -> 24.06 MiB | hist:
|
| 297 |
+
[ 267/ 291] blk.29.ffn_norm.weight - [ 4096, 1, 1, 1], type = f32, size = 0.016 MB
|
| 298 |
+
[ 268/ 291] blk.29.attn_k.weight - [ 4096, 1024, 1, 1], type = f32, quantizing to q3_K .. size = 16.00 MiB -> 1.72 MiB | hist:
|
| 299 |
+
[ 269/ 291] blk.29.attn_output.weight - [ 4096, 4096, 1, 1], type = f32, quantizing to q4_K .. size = 64.00 MiB -> 9.00 MiB | hist:
|
| 300 |
+
[ 270/ 291] blk.29.attn_q.weight - [ 4096, 4096, 1, 1], type = f32, quantizing to q3_K .. size = 64.00 MiB -> 6.88 MiB | hist:
|
| 301 |
+
[ 271/ 291] blk.29.attn_v.weight - [ 4096, 1024, 1, 1], type = f32, quantizing to q4_K .. size = 16.00 MiB -> 2.25 MiB | hist:
|
| 302 |
+
[ 272/ 291] blk.30.ffn_gate.weight - [ 4096, 14336, 1, 1], type = f32, quantizing to q3_K .. size = 224.00 MiB -> 24.06 MiB | hist:
|
| 303 |
+
[ 273/ 291] blk.30.ffn_up.weight - [ 4096, 14336, 1, 1], type = f32, quantizing to q3_K .. size = 224.00 MiB -> 24.06 MiB | hist:
|
| 304 |
+
[ 274/ 291] blk.30.attn_k.weight - [ 4096, 1024, 1, 1], type = f32, quantizing to q3_K .. size = 16.00 MiB -> 1.72 MiB | hist:
|
| 305 |
+
[ 275/ 291] blk.30.attn_output.weight - [ 4096, 4096, 1, 1], type = f32, quantizing to q4_K .. size = 64.00 MiB -> 9.00 MiB | hist:
|
| 306 |
+
[ 276/ 291] blk.30.attn_q.weight - [ 4096, 4096, 1, 1], type = f32, quantizing to q3_K .. size = 64.00 MiB -> 6.88 MiB | hist:
|
| 307 |
+
[ 277/ 291] blk.30.attn_v.weight - [ 4096, 1024, 1, 1], type = f32, quantizing to q4_K .. size = 16.00 MiB -> 2.25 MiB | hist:
|
| 308 |
+
[ 278/ 291] output.weight - [ 4096, 32000, 1, 1], type = f32, quantizing to q6_K .. size = 500.00 MiB -> 102.54 MiB | hist:
|
| 309 |
+
[ 279/ 291] blk.30.attn_norm.weight - [ 4096, 1, 1, 1], type = f32, size = 0.016 MB
|
| 310 |
+
[ 280/ 291] blk.30.ffn_down.weight - [14336, 4096, 1, 1], type = f32, quantizing to q4_K .. size = 224.00 MiB -> 31.50 MiB | hist:
|
| 311 |
+
[ 281/ 291] blk.30.ffn_norm.weight - [ 4096, 1, 1, 1], type = f32, size = 0.016 MB
|
| 312 |
+
[ 282/ 291] blk.31.attn_norm.weight - [ 4096, 1, 1, 1], type = f32, size = 0.016 MB
|
| 313 |
+
[ 283/ 291] blk.31.ffn_down.weight - [14336, 4096, 1, 1], type = f32, quantizing to q4_K .. size = 224.00 MiB -> 31.50 MiB | hist:
|
| 314 |
+
[ 284/ 291] blk.31.ffn_gate.weight - [ 4096, 14336, 1, 1], type = f32, quantizing to q3_K .. size = 224.00 MiB -> 24.06 MiB | hist:
|
| 315 |
+
[ 285/ 291] blk.31.ffn_up.weight - [ 4096, 14336, 1, 1], type = f32, quantizing to q3_K .. size = 224.00 MiB -> 24.06 MiB | hist:
|
| 316 |
+
[ 286/ 291] blk.31.ffn_norm.weight - [ 4096, 1, 1, 1], type = f32, size = 0.016 MB
|
| 317 |
+
[ 287/ 291] blk.31.attn_k.weight - [ 4096, 1024, 1, 1], type = f32, quantizing to q3_K .. size = 16.00 MiB -> 1.72 MiB | hist:
|
| 318 |
+
[ 288/ 291] blk.31.attn_output.weight - [ 4096, 4096, 1, 1], type = f32, quantizing to q4_K .. size = 64.00 MiB -> 9.00 MiB | hist:
|
| 319 |
+
[ 289/ 291] blk.31.attn_q.weight - [ 4096, 4096, 1, 1], type = f32, quantizing to q3_K .. size = 64.00 MiB -> 6.88 MiB | hist:
|
| 320 |
+
[ 290/ 291] blk.31.attn_v.weight - [ 4096, 1024, 1, 1], type = f32, quantizing to q4_K .. size = 16.00 MiB -> 2.25 MiB | hist:
|
| 321 |
+
[ 291/ 291] output_norm.weight - [ 4096, 1, 1, 1], type = f32, size = 0.016 MB
|
| 322 |
+
llama_model_quantize_internal: model size = 27625.02 MB
|
| 323 |
+
llama_model_quantize_internal: quant size = 3355.27 MB
|
| 324 |
+
|
| 325 |
+
main: quantize time = 368785.09 ms
|
| 326 |
+
main: total time = 368785.09 ms
|
zephyr_int8_int3.txt
ADDED
|
@@ -0,0 +1,362 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
(dream) tb@IBM-PF38WZKF:~/funstreams/AI$ ./llama.cpp/quantize zephyr_int8.gguf zephyr_Q3_K_M.gguf Q3_K_M
|
| 2 |
+
main: build = 1798 (128de35)
|
| 3 |
+
main: built with cc (Ubuntu 9.4.0-1ubuntu1~20.04.2) 9.4.0 for x86_64-linux-gnu
|
| 4 |
+
main: quantizing 'zephyr_int8.gguf' to 'zephyr_Q3_K_M.gguf' as Q3_K_M
|
| 5 |
+
llama_model_loader: loaded meta data with 22 key-value pairs and 291 tensors from zephyr_int8.gguf (version GGUF V3 (latest))
|
| 6 |
+
llama_model_loader: Dumping metadata keys/values. Note: KV overrides do not apply in this output.
|
| 7 |
+
llama_model_loader: - kv 0: general.architecture str = llama
|
| 8 |
+
llama_model_loader: - kv 1: general.name str = .
|
| 9 |
+
llama_model_loader: - kv 2: llama.context_length u32 = 32768
|
| 10 |
+
llama_model_loader: - kv 3: llama.embedding_length u32 = 4096
|
| 11 |
+
llama_model_loader: - kv 4: llama.block_count u32 = 32
|
| 12 |
+
llama_model_loader: - kv 5: llama.feed_forward_length u32 = 14336
|
| 13 |
+
llama_model_loader: - kv 6: llama.rope.dimension_count u32 = 128
|
| 14 |
+
llama_model_loader: - kv 7: llama.attention.head_count u32 = 32
|
| 15 |
+
llama_model_loader: - kv 8: llama.attention.head_count_kv u32 = 8
|
| 16 |
+
llama_model_loader: - kv 9: llama.attention.layer_norm_rms_epsilon f32 = 0.000010
|
| 17 |
+
llama_model_loader: - kv 10: llama.rope.freq_base f32 = 10000.000000
|
| 18 |
+
llama_model_loader: - kv 11: general.file_type u32 = 7
|
| 19 |
+
llama_model_loader: - kv 12: tokenizer.ggml.model str = llama
|
| 20 |
+
llama_model_loader: - kv 13: tokenizer.ggml.tokens arr[str,32000] = ["<unk>", "<s>", "</s>", "<0x00>", "<...
|
| 21 |
+
llama_model_loader: - kv 14: tokenizer.ggml.scores arr[f32,32000] = [0.000000, 0.000000, 0.000000, 0.0000...
|
| 22 |
+
llama_model_loader: - kv 15: tokenizer.ggml.token_type arr[i32,32000] = [2, 3, 3, 6, 6, 6, 6, 6, 6, 6, 6, 6, ...
|
| 23 |
+
llama_model_loader: - kv 16: tokenizer.ggml.merges arr[str,58980] = ["▁ t", "i n", "e r", "▁ a", "h e...
|
| 24 |
+
llama_model_loader: - kv 17: tokenizer.ggml.bos_token_id u32 = 1
|
| 25 |
+
llama_model_loader: - kv 18: tokenizer.ggml.eos_token_id u32 = 2
|
| 26 |
+
llama_model_loader: - kv 19: tokenizer.ggml.unknown_token_id u32 = 0
|
| 27 |
+
llama_model_loader: - kv 20: tokenizer.ggml.padding_token_id u32 = 2
|
| 28 |
+
llama_model_loader: - kv 21: tokenizer.chat_template str = {% for message in messages %}\n{% if m...
|
| 29 |
+
llama_model_loader: - type f32: 65 tensors
|
| 30 |
+
llama_model_loader: - type q8_0: 226 tensors
|
| 31 |
+
llama_model_quantize_internal: meta size = 1671648 bytes
|
| 32 |
+
[ 1/ 291] token_embd.weight - [ 4096, 32000, 1, 1], type = q8_0, llama_model_quantize: failed to quantize: requantizing from type q8_0 is disabled
|
| 33 |
+
main: failed to quantize model from 'zephyr_int8.gguf'
|
| 34 |
+
(dream) tb@IBM-PF38WZKF:~/funstreams/AI$ ./llama.cpp/quantize --allow-requantize zephyr_int8.gguf zephyr_Q3_K_M.gguf Q3_
|
| 35 |
+
K_M
|
| 36 |
+
main: build = 1798 (128de35)
|
| 37 |
+
main: built with cc (Ubuntu 9.4.0-1ubuntu1~20.04.2) 9.4.0 for x86_64-linux-gnu
|
| 38 |
+
main: quantizing 'zephyr_int8.gguf' to 'zephyr_Q3_K_M.gguf' as Q3_K_M
|
| 39 |
+
llama_model_loader: loaded meta data with 22 key-value pairs and 291 tensors from zephyr_int8.gguf (version GGUF V3 (latest))
|
| 40 |
+
llama_model_loader: Dumping metadata keys/values. Note: KV overrides do not apply in this output.
|
| 41 |
+
llama_model_loader: - kv 0: general.architecture str = llama
|
| 42 |
+
llama_model_loader: - kv 1: general.name str = .
|
| 43 |
+
llama_model_loader: - kv 2: llama.context_length u32 = 32768
|
| 44 |
+
llama_model_loader: - kv 3: llama.embedding_length u32 = 4096
|
| 45 |
+
llama_model_loader: - kv 4: llama.block_count u32 = 32
|
| 46 |
+
llama_model_loader: - kv 5: llama.feed_forward_length u32 = 14336
|
| 47 |
+
llama_model_loader: - kv 6: llama.rope.dimension_count u32 = 128
|
| 48 |
+
llama_model_loader: - kv 7: llama.attention.head_count u32 = 32
|
| 49 |
+
llama_model_loader: - kv 8: llama.attention.head_count_kv u32 = 8
|
| 50 |
+
llama_model_loader: - kv 9: llama.attention.layer_norm_rms_epsilon f32 = 0.000010
|
| 51 |
+
llama_model_loader: - kv 10: llama.rope.freq_base f32 = 10000.000000
|
| 52 |
+
llama_model_loader: - kv 11: general.file_type u32 = 7
|
| 53 |
+
llama_model_loader: - kv 12: tokenizer.ggml.model str = llama
|
| 54 |
+
llama_model_loader: - kv 13: tokenizer.ggml.tokens arr[str,32000] = ["<unk>", "<s>", "</s>", "<0x00>", "<...
|
| 55 |
+
llama_model_loader: - kv 14: tokenizer.ggml.scores arr[f32,32000] = [0.000000, 0.000000, 0.000000, 0.0000...
|
| 56 |
+
llama_model_loader: - kv 15: tokenizer.ggml.token_type arr[i32,32000] = [2, 3, 3, 6, 6, 6, 6, 6, 6, 6, 6, 6, ...
|
| 57 |
+
llama_model_loader: - kv 16: tokenizer.ggml.merges arr[str,58980] = ["▁ t", "i n", "e r", "▁ a", "h e...
|
| 58 |
+
llama_model_loader: - kv 17: tokenizer.ggml.bos_token_id u32 = 1
|
| 59 |
+
llama_model_loader: - kv 18: tokenizer.ggml.eos_token_id u32 = 2
|
| 60 |
+
llama_model_loader: - kv 19: tokenizer.ggml.unknown_token_id u32 = 0
|
| 61 |
+
llama_model_loader: - kv 20: tokenizer.ggml.padding_token_id u32 = 2
|
| 62 |
+
llama_model_loader: - kv 21: tokenizer.chat_template str = {% for message in messages %}\n{% if m...
|
| 63 |
+
llama_model_loader: - type f32: 65 tensors
|
| 64 |
+
llama_model_loader: - type q8_0: 226 tensors
|
| 65 |
+
llama_model_quantize_internal: meta size = 1671648 bytes
|
| 66 |
+
[ 1/ 291] token_embd.weight - [ 4096, 32000, 1, 1], type = q8_0, quantizing to q3_K .. size = 132.81 MiB -> 53.71 MiB | hist:
|
| 67 |
+
[ 2/ 291] blk.0.attn_norm.weight - [ 4096, 1, 1, 1], type = f32, size = 0.016 MB
|
| 68 |
+
[ 3/ 291] blk.0.ffn_down.weight - [14336, 4096, 1, 1], type = q8_0, quantizing to q5_K .. size = 59.50 MiB -> 38.50 MiB | hist:
|
| 69 |
+
[ 4/ 291] blk.0.ffn_gate.weight - [ 4096, 14336, 1, 1], type = q8_0, quantizing to q3_K .. size = 59.50 MiB -> 24.06 MiB | hist:
|
| 70 |
+
[ 5/ 291] blk.0.ffn_up.weight - [ 4096, 14336, 1, 1], type = q8_0, quantizing to q3_K .. size = 59.50 MiB -> 24.06 MiB | hist:
|
| 71 |
+
[ 6/ 291] blk.0.ffn_norm.weight - [ 4096, 1, 1, 1], type = f32, size = 0.016 MB
|
| 72 |
+
[ 7/ 291] blk.0.attn_k.weight - [ 4096, 1024, 1, 1], type = q8_0, quantizing to q3_K .. size = 4.25 MiB -> 1.72 MiB | hist:
|
| 73 |
+
[ 8/ 291] blk.0.attn_output.weight - [ 4096, 4096, 1, 1], type = q8_0, quantizing to q4_K .. size = 17.00 MiB -> 9.00 MiB | hist:
|
| 74 |
+
[ 9/ 291] blk.0.attn_q.weight - [ 4096, 4096, 1, 1], type = q8_0, quantizing to q3_K .. size = 17.00 MiB -> 6.88 MiB | hist:
|
| 75 |
+
[ 10/ 291] blk.0.attn_v.weight - [ 4096, 1024, 1, 1], type = q8_0, quantizing to q5_K .. size = 4.25 MiB -> 2.75 MiB | hist:
|
| 76 |
+
[ 11/ 291] blk.1.attn_norm.weight - [ 4096, 1, 1, 1], type = f32, size = 0.016 MB
|
| 77 |
+
[ 12/ 291] blk.1.ffn_down.weight - [14336, 4096, 1, 1], type = q8_0, quantizing to q5_K .. size = 59.50 MiB -> 38.50 MiB | hist:
|
| 78 |
+
[ 13/ 291] blk.1.ffn_gate.weight - [ 4096, 14336, 1, 1], type = q8_0, quantizing to q3_K .. size = 59.50 MiB -> 24.06 MiB | hist:
|
| 79 |
+
[ 14/ 291] blk.1.ffn_up.weight - [ 4096, 14336, 1, 1], type = q8_0, quantizing to q3_K .. size = 59.50 MiB -> 24.06 MiB | hist:
|
| 80 |
+
[ 15/ 291] blk.1.ffn_norm.weight - [ 4096, 1, 1, 1], type = f32, size = 0.016 MB
|
| 81 |
+
[ 16/ 291] blk.1.attn_k.weight - [ 4096, 1024, 1, 1], type = q8_0, quantizing to q3_K .. size = 4.25 MiB -> 1.72 MiB | hist:
|
| 82 |
+
[ 17/ 291] blk.1.attn_output.weight - [ 4096, 4096, 1, 1], type = q8_0, quantizing to q4_K .. size = 17.00 MiB -> 9.00 MiB | hist:
|
| 83 |
+
[ 18/ 291] blk.1.attn_q.weight - [ 4096, 4096, 1, 1], type = q8_0, quantizing to q3_K .. size = 17.00 MiB -> 6.88 MiB | hist:
|
| 84 |
+
[ 19/ 291] blk.1.attn_v.weight - [ 4096, 1024, 1, 1], type = q8_0, quantizing to q5_K .. size = 4.25 MiB -> 2.75 MiB | hist:
|
| 85 |
+
[ 20/ 291] blk.2.attn_norm.weight - [ 4096, 1, 1, 1], type = f32, size = 0.016 MB
|
| 86 |
+
[ 21/ 291] blk.2.ffn_down.weight - [14336, 4096, 1, 1], type = q8_0, quantizing to q4_K .. size = 59.50 MiB -> 31.50 MiB | hist:
|
| 87 |
+
[ 22/ 291] blk.2.ffn_gate.weight - [ 4096, 14336, 1, 1], type = q8_0, quantizing to q3_K .. size = 59.50 MiB -> 24.06 MiB | hist:
|
| 88 |
+
[ 23/ 291] blk.2.ffn_up.weight - [ 4096, 14336, 1, 1], type = q8_0, quantizing to q3_K .. size = 59.50 MiB -> 24.06 MiB | hist:
|
| 89 |
+
[ 24/ 291] blk.2.ffn_norm.weight - [ 4096, 1, 1, 1], type = f32, size = 0.016 MB
|
| 90 |
+
[ 25/ 291] blk.2.attn_k.weight - [ 4096, 1024, 1, 1], type = q8_0, quantizing to q3_K .. size = 4.25 MiB -> 1.72 MiB | hist:
|
| 91 |
+
[ 26/ 291] blk.2.attn_output.weight - [ 4096, 4096, 1, 1], type = q8_0, quantizing to q4_K .. size = 17.00 MiB -> 9.00 MiB | hist:
|
| 92 |
+
[ 27/ 291] blk.2.attn_q.weight - [ 4096, 4096, 1, 1], type = q8_0, quantizing to q3_K .. size = 17.00 MiB -> 6.88 MiB | hist:
|
| 93 |
+
[ 28/ 291] blk.2.attn_v.weight - [ 4096, 1024, 1, 1], type = q8_0, quantizing to q4_K .. size = 4.25 MiB -> 2.25 MiB | hist:
|
| 94 |
+
[ 29/ 291] blk.3.ffn_gate.weight - [ 4096, 14336, 1, 1], type = q8_0, quantizing to q3_K .. size = 59.50 MiB -> 24.06 MiB | hist:
|
| 95 |
+
[ 30/ 291] blk.3.ffn_up.weight - [ 4096, 14336, 1, 1], type = q8_0, quantizing to q3_K .. size = 59.50 MiB -> 24.06 MiB | hist:
|
| 96 |
+
[ 31/ 291] blk.3.attn_k.weight - [ 4096, 1024, 1, 1], type = q8_0, quantizing to q3_K .. size = 4.25 MiB -> 1.72 MiB | hist:
|
| 97 |
+
[ 32/ 291] blk.3.attn_output.weight - [ 4096, 4096, 1, 1], type = q8_0, quantizing to q4_K .. size = 17.00 MiB -> 9.00 MiB | hist:
|
| 98 |
+
[ 33/ 291] blk.3.attn_q.weight - [ 4096, 4096, 1, 1], type = q8_0, quantizing to q3_K .. size = 17.00 MiB -> 6.88 MiB | hist:
|
| 99 |
+
[ 34/ 291] blk.3.attn_v.weight - [ 4096, 1024, 1, 1], type = q8_0, quantizing to q4_K .. size = 4.25 MiB -> 2.25 MiB | hist:
|
| 100 |
+
[ 35/ 291] blk.3.attn_norm.weight - [ 4096, 1, 1, 1], type = f32, size = 0.016 MB
|
| 101 |
+
[ 36/ 291] blk.3.ffn_down.weight - [14336, 4096, 1, 1], type = q8_0, quantizing to q4_K .. size = 59.50 MiB -> 31.50 MiB | hist:
|
| 102 |
+
[ 37/ 291] blk.3.ffn_norm.weight - [ 4096, 1, 1, 1], type = f32, size = 0.016 MB
|
| 103 |
+
[ 38/ 291] blk.4.attn_norm.weight - [ 4096, 1, 1, 1], type = f32, size = 0.016 MB
|
| 104 |
+
[ 39/ 291] blk.4.ffn_down.weight - [14336, 4096, 1, 1], type = q8_0, quantizing to q4_K .. size = 59.50 MiB -> 31.50 MiB | hist:
|
| 105 |
+
[ 40/ 291] blk.4.ffn_gate.weight - [ 4096, 14336, 1, 1], type = q8_0, quantizing to q3_K .. size = 59.50 MiB -> 24.06 MiB | hist:
|
| 106 |
+
[ 41/ 291] blk.4.ffn_up.weight - [ 4096, 14336, 1, 1], type = q8_0, quantizing to q3_K .. size = 59.50 MiB -> 24.06 MiB | hist:
|
| 107 |
+
[ 42/ 291] blk.4.ffn_norm.weight - [ 4096, 1, 1, 1], type = f32, size = 0.016 MB
|
| 108 |
+
[ 43/ 291] blk.4.attn_k.weight - [ 4096, 1024, 1, 1], type = q8_0, quantizing to q3_K .. size = 4.25 MiB -> 1.72 MiB | hist:
|
| 109 |
+
[ 44/ 291] blk.4.attn_output.weight - [ 4096, 4096, 1, 1], type = q8_0, quantizing to q4_K .. size = 17.00 MiB -> 9.00 MiB | hist:
|
| 110 |
+
[ 45/ 291] blk.4.attn_q.weight - [ 4096, 4096, 1, 1], type = q8_0, quantizing to q3_K .. size = 17.00 MiB -> 6.88 MiB | hist:
|
| 111 |
+
[ 46/ 291] blk.4.attn_v.weight - [ 4096, 1024, 1, 1], type = q8_0, quantizing to q4_K .. size = 4.25 MiB -> 2.25 MiB | hist:
|
| 112 |
+
[ 47/ 291] blk.5.attn_norm.weight - [ 4096, 1, 1, 1], type = f32, size = 0.016 MB
|
| 113 |
+
[ 48/ 291] blk.5.ffn_down.weight - [14336, 4096, 1, 1], type = q8_0, quantizing to q4_K .. size = 59.50 MiB -> 31.50 MiB | hist:
|
| 114 |
+
[ 49/ 291] blk.5.ffn_gate.weight - [ 4096, 14336, 1, 1], type = q8_0, quantizing to q3_K .. size = 59.50 MiB -> 24.06 MiB | hist:
|
| 115 |
+
[ 50/ 291] blk.5.ffn_up.weight - [ 4096, 14336, 1, 1], type = q8_0, quantizing to q3_K .. size = 59.50 MiB -> 24.06 MiB | hist:
|
| 116 |
+
[ 51/ 291] blk.5.ffn_norm.weight - [ 4096, 1, 1, 1], type = f32, size = 0.016 MB
|
| 117 |
+
[ 52/ 291] blk.5.attn_k.weight - [ 4096, 1024, 1, 1], type = q8_0, quantizing to q3_K .. size = 4.25 MiB -> 1.72 MiB | hist:
|
| 118 |
+
[ 53/ 291] blk.5.attn_output.weight - [ 4096, 4096, 1, 1], type = q8_0, quantizing to q4_K .. size = 17.00 MiB -> 9.00 MiB | hist:
|
| 119 |
+
[ 54/ 291] blk.5.attn_q.weight - [ 4096, 4096, 1, 1], type = q8_0, quantizing to q3_K .. size = 17.00 MiB -> 6.88 MiB | hist:
|
| 120 |
+
[ 55/ 291] blk.5.attn_v.weight - [ 4096, 1024, 1, 1], type = q8_0, quantizing to q4_K .. size = 4.25 MiB -> 2.25 MiB | hist:
|
| 121 |
+
[ 56/ 291] blk.6.attn_norm.weight - [ 4096, 1, 1, 1], type = f32, size = 0.016 MB
|
| 122 |
+
[ 57/ 291] blk.6.ffn_down.weight - [14336, 4096, 1, 1], type = q8_0, quantizing to q4_K .. size = 59.50 MiB -> 31.50 MiB | hist:
|
| 123 |
+
[ 58/ 291] blk.6.ffn_gate.weight - [ 4096, 14336, 1, 1], type = q8_0, quantizing to q3_K .. size = 59.50 MiB -> 24.06 MiB | hist:
|
| 124 |
+
[ 59/ 291] blk.6.ffn_up.weight - [ 4096, 14336, 1, 1], type = q8_0, quantizing to q3_K .. size = 59.50 MiB -> 24.06 MiB | hist:
|
| 125 |
+
[ 60/ 291] blk.6.ffn_norm.weight - [ 4096, 1, 1, 1], type = f32, size = 0.016 MB
|
| 126 |
+
[ 61/ 291] blk.6.attn_k.weight - [ 4096, 1024, 1, 1], type = q8_0, quantizing to q3_K .. size = 4.25 MiB -> 1.72 MiB | hist:
|
| 127 |
+
[ 62/ 291] blk.6.attn_output.weight - [ 4096, 4096, 1, 1], type = q8_0, quantizing to q4_K .. size = 17.00 MiB -> 9.00 MiB | hist:
|
| 128 |
+
[ 63/ 291] blk.6.attn_q.weight - [ 4096, 4096, 1, 1], type = q8_0, quantizing to q3_K .. size = 17.00 MiB -> 6.88 MiB | hist:
|
| 129 |
+
[ 64/ 291] blk.6.attn_v.weight - [ 4096, 1024, 1, 1], type = q8_0, quantizing to q4_K .. size = 4.25 MiB -> 2.25 MiB | hist:
|
| 130 |
+
[ 65/ 291] blk.7.attn_norm.weight - [ 4096, 1, 1, 1], type = f32, size = 0.016 MB
|
| 131 |
+
[ 66/ 291] blk.7.ffn_down.weight - [14336, 4096, 1, 1], type = q8_0, quantizing to q4_K .. size = 59.50 MiB -> 31.50 MiB | hist:
|
| 132 |
+
[ 67/ 291] blk.7.ffn_gate.weight - [ 4096, 14336, 1, 1], type = q8_0, quantizing to q3_K .. size = 59.50 MiB -> 24.06 MiB | hist:
|
| 133 |
+
[ 68/ 291] blk.7.ffn_up.weight - [ 4096, 14336, 1, 1], type = q8_0, quantizing to q3_K .. size = 59.50 MiB -> 24.06 MiB | hist:
|
| 134 |
+
[ 69/ 291] blk.7.ffn_norm.weight - [ 4096, 1, 1, 1], type = f32, size = 0.016 MB
|
| 135 |
+
[ 70/ 291] blk.7.attn_k.weight - [ 4096, 1024, 1, 1], type = q8_0, quantizing to q3_K .. size = 4.25 MiB -> 1.72 MiB | hist:
|
| 136 |
+
[ 71/ 291] blk.7.attn_output.weight - [ 4096, 4096, 1, 1], type = q8_0, quantizing to q4_K .. size = 17.00 MiB -> 9.00 MiB | hist:
|
| 137 |
+
[ 72/ 291] blk.7.attn_q.weight - [ 4096, 4096, 1, 1], type = q8_0, quantizing to q3_K .. size = 17.00 MiB -> 6.88 MiB | hist:
|
| 138 |
+
[ 73/ 291] blk.7.attn_v.weight - [ 4096, 1024, 1, 1], type = q8_0, quantizing to q4_K .. size = 4.25 MiB -> 2.25 MiB | hist:
|
| 139 |
+
[ 74/ 291] blk.8.attn_k.weight - [ 4096, 1024, 1, 1], type = q8_0, quantizing to q3_K .. size = 4.25 MiB -> 1.72 MiB | hist:
|
| 140 |
+
[ 75/ 291] blk.8.attn_output.weight - [ 4096, 4096, 1, 1], type = q8_0, quantizing to q4_K .. size = 17.00 MiB -> 9.00 MiB | hist:
|
| 141 |
+
[ 76/ 291] blk.8.attn_q.weight - [ 4096, 4096, 1, 1], type = q8_0, quantizing to q3_K .. size = 17.00 MiB -> 6.88 MiB | hist:
|
| 142 |
+
[ 77/ 291] blk.8.attn_v.weight - [ 4096, 1024, 1, 1], type = q8_0, quantizing to q4_K .. size = 4.25 MiB -> 2.25 MiB | hist:
|
| 143 |
+
[ 78/ 291] blk.10.attn_norm.weight - [ 4096, 1, 1, 1], type = f32, size = 0.016 MB
|
| 144 |
+
[ 79/ 291] blk.10.ffn_down.weight - [14336, 4096, 1, 1], type = q8_0, quantizing to q4_K .. size = 59.50 MiB -> 31.50 MiB | hist:
|
| 145 |
+
[ 80/ 291] blk.10.ffn_gate.weight - [ 4096, 14336, 1, 1], type = q8_0, quantizing to q3_K .. size = 59.50 MiB -> 24.06 MiB | hist:
|
| 146 |
+
[ 81/ 291] blk.10.ffn_up.weight - [ 4096, 14336, 1, 1], type = q8_0, quantizing to q3_K .. size = 59.50 MiB -> 24.06 MiB | hist:
|
| 147 |
+
[ 82/ 291] blk.10.ffn_norm.weight - [ 4096, 1, 1, 1], type = f32, size = 0.016 MB
|
| 148 |
+
[ 83/ 291] blk.10.attn_k.weight - [ 4096, 1024, 1, 1], type = q8_0, quantizing to q3_K .. size = 4.25 MiB -> 1.72 MiB | hist:
|
| 149 |
+
[ 84/ 291] blk.10.attn_output.weight - [ 4096, 4096, 1, 1], type = q8_0, quantizing to q4_K .. size = 17.00 MiB -> 9.00 MiB | hist:
|
| 150 |
+
[ 85/ 291] blk.10.attn_q.weight - [ 4096, 4096, 1, 1], type = q8_0, quantizing to q3_K .. size = 17.00 MiB -> 6.88 MiB | hist:
|
| 151 |
+
[ 86/ 291] blk.10.attn_v.weight - [ 4096, 1024, 1, 1], type = q8_0, quantizing to q4_K .. size = 4.25 MiB -> 2.25 MiB | hist:
|
| 152 |
+
[ 87/ 291] blk.11.attn_norm.weight - [ 4096, 1, 1, 1], type = f32, size = 0.016 MB
|
| 153 |
+
[ 88/ 291] blk.11.ffn_down.weight - [14336, 4096, 1, 1], type = q8_0, quantizing to q4_K .. size = 59.50 MiB -> 31.50 MiB | hist:
|
| 154 |
+
[ 89/ 291] blk.11.ffn_gate.weight - [ 4096, 14336, 1, 1], type = q8_0, quantizing to q3_K .. size = 59.50 MiB -> 24.06 MiB | hist:
|
| 155 |
+
[ 90/ 291] blk.11.ffn_up.weight - [ 4096, 14336, 1, 1], type = q8_0, quantizing to q3_K .. size = 59.50 MiB -> 24.06 MiB | hist:
|
| 156 |
+
[ 91/ 291] blk.11.ffn_norm.weight - [ 4096, 1, 1, 1], type = f32, size = 0.016 MB
|
| 157 |
+
[ 92/ 291] blk.11.attn_k.weight - [ 4096, 1024, 1, 1], type = q8_0, quantizing to q3_K .. size = 4.25 MiB -> 1.72 MiB | hist:
|
| 158 |
+
[ 93/ 291] blk.11.attn_output.weight - [ 4096, 4096, 1, 1], type = q8_0, quantizing to q4_K .. size = 17.00 MiB -> 9.00 MiB | hist:
|
| 159 |
+
[ 94/ 291] blk.11.attn_q.weight - [ 4096, 4096, 1, 1], type = q8_0, quantizing to q3_K .. size = 17.00 MiB -> 6.88 MiB | hist:
|
| 160 |
+
[ 95/ 291] blk.11.attn_v.weight - [ 4096, 1024, 1, 1], type = q8_0, quantizing to q4_K .. size = 4.25 MiB -> 2.25 MiB | hist:
|
| 161 |
+
[ 96/ 291] blk.12.ffn_gate.weight - [ 4096, 14336, 1, 1], type = q8_0, quantizing to q3_K .. size = 59.50 MiB -> 24.06 MiB | hist:
|
| 162 |
+
[ 97/ 291] blk.12.ffn_up.weight - [ 4096, 14336, 1, 1], type = q8_0, quantizing to q3_K .. size = 59.50 MiB -> 24.06 MiB | hist:
|
| 163 |
+
[ 98/ 291] blk.12.attn_k.weight - [ 4096, 1024, 1, 1], type = q8_0, quantizing to q3_K .. size = 4.25 MiB -> 1.72 MiB | hist:
|
| 164 |
+
[ 99/ 291] blk.12.attn_output.weight - [ 4096, 4096, 1, 1], type = q8_0, quantizing to q4_K .. size = 17.00 MiB -> 9.00 MiB | hist:
|
| 165 |
+
[ 100/ 291] blk.12.attn_q.weight - [ 4096, 4096, 1, 1], type = q8_0, quantizing to q3_K .. size = 17.00 MiB -> 6.88 MiB | hist:
|
| 166 |
+
[ 101/ 291] blk.12.attn_v.weight - [ 4096, 1024, 1, 1], type = q8_0, quantizing to q4_K .. size = 4.25 MiB -> 2.25 MiB | hist:
|
| 167 |
+
[ 102/ 291] blk.8.attn_norm.weight - [ 4096, 1, 1, 1], type = f32, size = 0.016 MB
|
| 168 |
+
[ 103/ 291] blk.8.ffn_down.weight - [14336, 4096, 1, 1], type = q8_0, quantizing to q4_K .. size = 59.50 MiB -> 31.50 MiB | hist:
|
| 169 |
+
[ 104/ 291] blk.8.ffn_gate.weight - [ 4096, 14336, 1, 1], type = q8_0, quantizing to q3_K .. size = 59.50 MiB -> 24.06 MiB | hist:
|
| 170 |
+
[ 105/ 291] blk.8.ffn_up.weight - [ 4096, 14336, 1, 1], type = q8_0, quantizing to q3_K .. size = 59.50 MiB -> 24.06 MiB | hist:
|
| 171 |
+
[ 106/ 291] blk.8.ffn_norm.weight - [ 4096, 1, 1, 1], type = f32, size = 0.016 MB
|
| 172 |
+
[ 107/ 291] blk.9.attn_norm.weight - [ 4096, 1, 1, 1], type = f32, size = 0.016 MB
|
| 173 |
+
[ 108/ 291] blk.9.ffn_down.weight - [14336, 4096, 1, 1], type = q8_0, quantizing to q4_K .. size = 59.50 MiB -> 31.50 MiB | hist:
|
| 174 |
+
[ 109/ 291] blk.9.ffn_gate.weight - [ 4096, 14336, 1, 1], type = q8_0, quantizing to q3_K .. size = 59.50 MiB -> 24.06 MiB | hist:
|
| 175 |
+
[ 110/ 291] blk.9.ffn_up.weight - [ 4096, 14336, 1, 1], type = q8_0, quantizing to q3_K .. size = 59.50 MiB -> 24.06 MiB | hist:
|
| 176 |
+
[ 111/ 291] blk.9.ffn_norm.weight - [ 4096, 1, 1, 1], type = f32, size = 0.016 MB
|
| 177 |
+
[ 112/ 291] blk.9.attn_k.weight - [ 4096, 1024, 1, 1], type = q8_0, quantizing to q3_K .. size = 4.25 MiB -> 1.72 MiB | hist:
|
| 178 |
+
[ 113/ 291] blk.9.attn_output.weight - [ 4096, 4096, 1, 1], type = q8_0, quantizing to q4_K .. size = 17.00 MiB -> 9.00 MiB | hist:
|
| 179 |
+
[ 114/ 291] blk.9.attn_q.weight - [ 4096, 4096, 1, 1], type = q8_0, quantizing to q3_K .. size = 17.00 MiB -> 6.88 MiB | hist:
|
| 180 |
+
[ 115/ 291] blk.9.attn_v.weight - [ 4096, 1024, 1, 1], type = q8_0, quantizing to q4_K .. size = 4.25 MiB -> 2.25 MiB | hist:
|
| 181 |
+
[ 116/ 291] blk.12.attn_norm.weight - [ 4096, 1, 1, 1], type = f32, size = 0.016 MB
|
| 182 |
+
[ 117/ 291] blk.12.ffn_down.weight - [14336, 4096, 1, 1], type = q8_0, quantizing to q4_K .. size = 59.50 MiB -> 31.50 MiB | hist:
|
| 183 |
+
[ 118/ 291] blk.12.ffn_norm.weight - [ 4096, 1, 1, 1], type = f32, size = 0.016 MB
|
| 184 |
+
[ 119/ 291] blk.13.attn_norm.weight - [ 4096, 1, 1, 1], type = f32, size = 0.016 MB
|
| 185 |
+
[ 120/ 291] blk.13.ffn_down.weight - [14336, 4096, 1, 1], type = q8_0, quantizing to q4_K .. size = 59.50 MiB -> 31.50 MiB | hist:
|
| 186 |
+
[ 121/ 291] blk.13.ffn_gate.weight - [ 4096, 14336, 1, 1], type = q8_0, quantizing to q3_K .. size = 59.50 MiB -> 24.06 MiB | hist:
|
| 187 |
+
[ 122/ 291] blk.13.ffn_up.weight - [ 4096, 14336, 1, 1], type = q8_0, quantizing to q3_K .. size = 59.50 MiB -> 24.06 MiB | hist:
|
| 188 |
+
[ 123/ 291] blk.13.ffn_norm.weight - [ 4096, 1, 1, 1], type = f32, size = 0.016 MB
|
| 189 |
+
[ 124/ 291] blk.13.attn_k.weight - [ 4096, 1024, 1, 1], type = q8_0, quantizing to q3_K .. size = 4.25 MiB -> 1.72 MiB | hist:
|
| 190 |
+
[ 125/ 291] blk.13.attn_output.weight - [ 4096, 4096, 1, 1], type = q8_0, quantizing to q4_K .. size = 17.00 MiB -> 9.00 MiB | hist:
|
| 191 |
+
[ 126/ 291] blk.13.attn_q.weight - [ 4096, 4096, 1, 1], type = q8_0, quantizing to q3_K .. size = 17.00 MiB -> 6.88 MiB | hist:
|
| 192 |
+
[ 127/ 291] blk.13.attn_v.weight - [ 4096, 1024, 1, 1], type = q8_0, quantizing to q4_K .. size = 4.25 MiB -> 2.25 MiB | hist:
|
| 193 |
+
[ 128/ 291] blk.14.attn_norm.weight - [ 4096, 1, 1, 1], type = f32, size = 0.016 MB
|
| 194 |
+
[ 129/ 291] blk.14.ffn_down.weight - [14336, 4096, 1, 1], type = q8_0, quantizing to q4_K .. size = 59.50 MiB -> 31.50 MiB | hist:
|
| 195 |
+
[ 130/ 291] blk.14.ffn_gate.weight - [ 4096, 14336, 1, 1], type = q8_0, quantizing to q3_K .. size = 59.50 MiB -> 24.06 MiB | hist:
|
| 196 |
+
[ 131/ 291] blk.14.ffn_up.weight - [ 4096, 14336, 1, 1], type = q8_0, quantizing to q3_K .. size = 59.50 MiB -> 24.06 MiB | hist:
|
| 197 |
+
[ 132/ 291] blk.14.ffn_norm.weight - [ 4096, 1, 1, 1], type = f32, size = 0.016 MB
|
| 198 |
+
[ 133/ 291] blk.14.attn_k.weight - [ 4096, 1024, 1, 1], type = q8_0, quantizing to q3_K .. size = 4.25 MiB -> 1.72 MiB | hist:
|
| 199 |
+
[ 134/ 291] blk.14.attn_output.weight - [ 4096, 4096, 1, 1], type = q8_0, quantizing to q4_K .. size = 17.00 MiB -> 9.00 MiB | hist:
|
| 200 |
+
[ 135/ 291] blk.14.attn_q.weight - [ 4096, 4096, 1, 1], type = q8_0, quantizing to q3_K .. size = 17.00 MiB -> 6.88 MiB | hist:
|
| 201 |
+
[ 136/ 291] blk.14.attn_v.weight - [ 4096, 1024, 1, 1], type = q8_0, quantizing to q4_K .. size = 4.25 MiB -> 2.25 MiB | hist:
|
| 202 |
+
[ 137/ 291] blk.15.attn_norm.weight - [ 4096, 1, 1, 1], type = f32, size = 0.016 MB
|
| 203 |
+
[ 138/ 291] blk.15.ffn_down.weight - [14336, 4096, 1, 1], type = q8_0, quantizing to q4_K .. size = 59.50 MiB -> 31.50 MiB | hist:
|
| 204 |
+
[ 139/ 291] blk.15.ffn_gate.weight - [ 4096, 14336, 1, 1], type = q8_0, quantizing to q3_K .. size = 59.50 MiB -> 24.06 MiB | hist:
|
| 205 |
+
[ 140/ 291] blk.15.ffn_up.weight - [ 4096, 14336, 1, 1], type = q8_0, quantizing to q3_K .. size = 59.50 MiB -> 24.06 MiB | hist:
|
| 206 |
+
[ 141/ 291] blk.15.ffn_norm.weight - [ 4096, 1, 1, 1], type = f32, size = 0.016 MB
|
| 207 |
+
[ 142/ 291] blk.15.attn_k.weight - [ 4096, 1024, 1, 1], type = q8_0, quantizing to q3_K .. size = 4.25 MiB -> 1.72 MiB | hist:
|
| 208 |
+
[ 143/ 291] blk.15.attn_output.weight - [ 4096, 4096, 1, 1], type = q8_0, quantizing to q4_K .. size = 17.00 MiB -> 9.00 MiB | hist:
|
| 209 |
+
[ 144/ 291] blk.15.attn_q.weight - [ 4096, 4096, 1, 1], type = q8_0, quantizing to q3_K .. size = 17.00 MiB -> 6.88 MiB | hist:
|
| 210 |
+
[ 145/ 291] blk.15.attn_v.weight - [ 4096, 1024, 1, 1], type = q8_0, quantizing to q4_K .. size = 4.25 MiB -> 2.25 MiB | hist:
|
| 211 |
+
[ 146/ 291] blk.16.attn_norm.weight - [ 4096, 1, 1, 1], type = f32, size = 0.016 MB
|
| 212 |
+
[ 147/ 291] blk.16.ffn_down.weight - [14336, 4096, 1, 1], type = q8_0, quantizing to q4_K .. size = 59.50 MiB -> 31.50 MiB | hist:
|
| 213 |
+
[ 148/ 291] blk.16.ffn_gate.weight - [ 4096, 14336, 1, 1], type = q8_0, quantizing to q3_K .. size = 59.50 MiB -> 24.06 MiB | hist:
|
| 214 |
+
[ 149/ 291] blk.16.ffn_up.weight - [ 4096, 14336, 1, 1], type = q8_0, quantizing to q3_K .. size = 59.50 MiB -> 24.06 MiB | hist:
|
| 215 |
+
[ 150/ 291] blk.16.ffn_norm.weight - [ 4096, 1, 1, 1], type = f32, size = 0.016 MB
|
| 216 |
+
[ 151/ 291] blk.16.attn_k.weight - [ 4096, 1024, 1, 1], type = q8_0, quantizing to q3_K .. size = 4.25 MiB -> 1.72 MiB | hist:
|
| 217 |
+
[ 152/ 291] blk.16.attn_output.weight - [ 4096, 4096, 1, 1], type = q8_0, quantizing to q4_K .. size = 17.00 MiB -> 9.00 MiB | hist:
|
| 218 |
+
[ 153/ 291] blk.16.attn_q.weight - [ 4096, 4096, 1, 1], type = q8_0, quantizing to q3_K .. size = 17.00 MiB -> 6.88 MiB | hist:
|
| 219 |
+
[ 154/ 291] blk.16.attn_v.weight - [ 4096, 1024, 1, 1], type = q8_0, quantizing to q4_K .. size = 4.25 MiB -> 2.25 MiB | hist:
|
| 220 |
+
[ 155/ 291] blk.17.attn_k.weight - [ 4096, 1024, 1, 1], type = q8_0, quantizing to q3_K .. size = 4.25 MiB -> 1.72 MiB | hist:
|
| 221 |
+
[ 156/ 291] blk.17.attn_output.weight - [ 4096, 4096, 1, 1], type = q8_0, quantizing to q4_K .. size = 17.00 MiB -> 9.00 MiB | hist:
|
| 222 |
+
[ 157/ 291] blk.17.attn_q.weight - [ 4096, 4096, 1, 1], type = q8_0, quantizing to q3_K .. size = 17.00 MiB -> 6.88 MiB | hist:
|
| 223 |
+
[ 158/ 291] blk.17.attn_v.weight - [ 4096, 1024, 1, 1], type = q8_0, quantizing to q4_K .. size = 4.25 MiB -> 2.25 MiB | hist:
|
| 224 |
+
[ 159/ 291] blk.17.attn_norm.weight - [ 4096, 1, 1, 1], type = f32, size = 0.016 MB
|
| 225 |
+
[ 160/ 291] blk.17.ffn_down.weight - [14336, 4096, 1, 1], type = q8_0, quantizing to q4_K .. size = 59.50 MiB -> 31.50 MiB | hist:
|
| 226 |
+
[ 161/ 291] blk.17.ffn_gate.weight - [ 4096, 14336, 1, 1], type = q8_0, quantizing to q3_K .. size = 59.50 MiB -> 24.06 MiB | hist:
|
| 227 |
+
[ 162/ 291] blk.17.ffn_up.weight - [ 4096, 14336, 1, 1], type = q8_0, quantizing to q3_K .. size = 59.50 MiB -> 24.06 MiB | hist:
|
| 228 |
+
[ 163/ 291] blk.17.ffn_norm.weight - [ 4096, 1, 1, 1], type = f32, size = 0.016 MB
|
| 229 |
+
[ 164/ 291] blk.18.attn_norm.weight - [ 4096, 1, 1, 1], type = f32, size = 0.016 MB
|
| 230 |
+
[ 165/ 291] blk.18.ffn_down.weight - [14336, 4096, 1, 1], type = q8_0, quantizing to q4_K .. size = 59.50 MiB -> 31.50 MiB | hist:
|
| 231 |
+
[ 166/ 291] blk.18.ffn_gate.weight - [ 4096, 14336, 1, 1], type = q8_0, quantizing to q3_K .. size = 59.50 MiB -> 24.06 MiB | hist:
|
| 232 |
+
[ 167/ 291] blk.18.ffn_up.weight - [ 4096, 14336, 1, 1], type = q8_0, quantizing to q3_K .. size = 59.50 MiB -> 24.06 MiB | hist:
|
| 233 |
+
[ 168/ 291] blk.18.ffn_norm.weight - [ 4096, 1, 1, 1], type = f32, size = 0.016 MB
|
| 234 |
+
[ 169/ 291] blk.18.attn_k.weight - [ 4096, 1024, 1, 1], type = q8_0, quantizing to q3_K .. size = 4.25 MiB -> 1.72 MiB | hist:
|
| 235 |
+
[ 170/ 291] blk.18.attn_output.weight - [ 4096, 4096, 1, 1], type = q8_0, quantizing to q4_K .. size = 17.00 MiB -> 9.00 MiB | hist:
|
| 236 |
+
[ 171/ 291] blk.18.attn_q.weight - [ 4096, 4096, 1, 1], type = q8_0, quantizing to q3_K .. size = 17.00 MiB -> 6.88 MiB | hist:
|
| 237 |
+
[ 172/ 291] blk.18.attn_v.weight - [ 4096, 1024, 1, 1], type = q8_0, quantizing to q4_K .. size = 4.25 MiB -> 2.25 MiB | hist:
|
| 238 |
+
[ 173/ 291] blk.19.attn_norm.weight - [ 4096, 1, 1, 1], type = f32, size = 0.016 MB
|
| 239 |
+
[ 174/ 291] blk.19.ffn_down.weight - [14336, 4096, 1, 1], type = q8_0, quantizing to q4_K .. size = 59.50 MiB -> 31.50 MiB | hist:
|
| 240 |
+
[ 175/ 291] blk.19.ffn_gate.weight - [ 4096, 14336, 1, 1], type = q8_0, quantizing to q3_K .. size = 59.50 MiB -> 24.06 MiB | hist:
|
| 241 |
+
[ 176/ 291] blk.19.ffn_up.weight - [ 4096, 14336, 1, 1], type = q8_0, quantizing to q3_K .. size = 59.50 MiB -> 24.06 MiB | hist:
|
| 242 |
+
[ 177/ 291] blk.19.ffn_norm.weight - [ 4096, 1, 1, 1], type = f32, size = 0.016 MB
|
| 243 |
+
[ 178/ 291] blk.19.attn_k.weight - [ 4096, 1024, 1, 1], type = q8_0, quantizing to q3_K .. size = 4.25 MiB -> 1.72 MiB | hist:
|
| 244 |
+
[ 179/ 291] blk.19.attn_output.weight - [ 4096, 4096, 1, 1], type = q8_0, quantizing to q4_K .. size = 17.00 MiB -> 9.00 MiB | hist:
|
| 245 |
+
[ 180/ 291] blk.19.attn_q.weight - [ 4096, 4096, 1, 1], type = q8_0, quantizing to q3_K .. size = 17.00 MiB -> 6.88 MiB | hist:
|
| 246 |
+
[ 181/ 291] blk.19.attn_v.weight - [ 4096, 1024, 1, 1], type = q8_0, quantizing to q4_K .. size = 4.25 MiB -> 2.25 MiB | hist:
|
| 247 |
+
[ 182/ 291] blk.20.attn_norm.weight - [ 4096, 1, 1, 1], type = f32, size = 0.016 MB
|
| 248 |
+
[ 183/ 291] blk.20.ffn_down.weight - [14336, 4096, 1, 1], type = q8_0, quantizing to q4_K .. size = 59.50 MiB -> 31.50 MiB | hist:
|
| 249 |
+
[ 184/ 291] blk.20.ffn_gate.weight - [ 4096, 14336, 1, 1], type = q8_0, quantizing to q3_K .. size = 59.50 MiB -> 24.06 MiB | hist:
|
| 250 |
+
[ 185/ 291] blk.20.ffn_up.weight - [ 4096, 14336, 1, 1], type = q8_0, quantizing to q3_K .. size = 59.50 MiB -> 24.06 MiB | hist:
|
| 251 |
+
[ 186/ 291] blk.20.ffn_norm.weight - [ 4096, 1, 1, 1], type = f32, size = 0.016 MB
|
| 252 |
+
[ 187/ 291] blk.20.attn_k.weight - [ 4096, 1024, 1, 1], type = q8_0, quantizing to q3_K .. size = 4.25 MiB -> 1.72 MiB | hist:
|
| 253 |
+
[ 188/ 291] blk.20.attn_output.weight - [ 4096, 4096, 1, 1], type = q8_0, quantizing to q4_K .. size = 17.00 MiB -> 9.00 MiB | hist:
|
| 254 |
+
[ 189/ 291] blk.20.attn_q.weight - [ 4096, 4096, 1, 1], type = q8_0, quantizing to q3_K .. size = 17.00 MiB -> 6.88 MiB | hist:
|
| 255 |
+
[ 190/ 291] blk.20.attn_v.weight - [ 4096, 1024, 1, 1], type = q8_0, quantizing to q4_K .. size = 4.25 MiB -> 2.25 MiB | hist:
|
| 256 |
+
[ 191/ 291] blk.21.ffn_gate.weight - [ 4096, 14336, 1, 1], type = q8_0, quantizing to q3_K .. size = 59.50 MiB -> 24.06 MiB | hist:
|
| 257 |
+
[ 192/ 291] blk.21.ffn_up.weight - [ 4096, 14336, 1, 1], type = q8_0, quantizing to q3_K .. size = 59.50 MiB -> 24.06 MiB | hist:
|
| 258 |
+
[ 193/ 291] blk.21.attn_k.weight - [ 4096, 1024, 1, 1], type = q8_0, quantizing to q3_K .. size = 4.25 MiB -> 1.72 MiB | hist:
|
| 259 |
+
[ 194/ 291] blk.21.attn_output.weight - [ 4096, 4096, 1, 1], type = q8_0, quantizing to q4_K .. size = 17.00 MiB -> 9.00 MiB | hist:
|
| 260 |
+
[ 195/ 291] blk.21.attn_q.weight - [ 4096, 4096, 1, 1], type = q8_0, quantizing to q3_K .. size = 17.00 MiB -> 6.88 MiB | hist:
|
| 261 |
+
[ 196/ 291] blk.21.attn_v.weight - [ 4096, 1024, 1, 1], type = q8_0, quantizing to q4_K .. size = 4.25 MiB -> 2.25 MiB | hist:
|
| 262 |
+
[ 197/ 291] blk.21.attn_norm.weight - [ 4096, 1, 1, 1], type = f32, size = 0.016 MB
|
| 263 |
+
[ 198/ 291] blk.21.ffn_down.weight - [14336, 4096, 1, 1], type = q8_0, quantizing to q4_K .. size = 59.50 MiB -> 31.50 MiB | hist:
|
| 264 |
+
[ 199/ 291] blk.21.ffn_norm.weight - [ 4096, 1, 1, 1], type = f32, size = 0.016 MB
|
| 265 |
+
[ 200/ 291] blk.22.attn_norm.weight - [ 4096, 1, 1, 1], type = f32, size = 0.016 MB
|
| 266 |
+
[ 201/ 291] blk.22.ffn_down.weight - [14336, 4096, 1, 1], type = q8_0, quantizing to q4_K .. size = 59.50 MiB -> 31.50 MiB | hist:
|
| 267 |
+
[ 202/ 291] blk.22.ffn_gate.weight - [ 4096, 14336, 1, 1], type = q8_0, quantizing to q3_K .. size = 59.50 MiB -> 24.06 MiB | hist:
|
| 268 |
+
[ 203/ 291] blk.22.ffn_up.weight - [ 4096, 14336, 1, 1], type = q8_0, quantizing to q3_K .. size = 59.50 MiB -> 24.06 MiB | hist:
|
| 269 |
+
[ 204/ 291] blk.22.ffn_norm.weight - [ 4096, 1, 1, 1], type = f32, size = 0.016 MB
|
| 270 |
+
[ 205/ 291] blk.22.attn_k.weight - [ 4096, 1024, 1, 1], type = q8_0, quantizing to q3_K .. size = 4.25 MiB -> 1.72 MiB | hist:
|
| 271 |
+
[ 206/ 291] blk.22.attn_output.weight - [ 4096, 4096, 1, 1], type = q8_0, quantizing to q4_K .. size = 17.00 MiB -> 9.00 MiB | hist:
|
| 272 |
+
[ 207/ 291] blk.22.attn_q.weight - [ 4096, 4096, 1, 1], type = q8_0, quantizing to q3_K .. size = 17.00 MiB -> 6.88 MiB | hist:
|
| 273 |
+
[ 208/ 291] blk.22.attn_v.weight - [ 4096, 1024, 1, 1], type = q8_0, quantizing to q4_K .. size = 4.25 MiB -> 2.25 MiB | hist:
|
| 274 |
+
[ 209/ 291] blk.23.attn_norm.weight - [ 4096, 1, 1, 1], type = f32, size = 0.016 MB
|
| 275 |
+
[ 210/ 291] blk.23.ffn_down.weight - [14336, 4096, 1, 1], type = q8_0, quantizing to q4_K .. size = 59.50 MiB -> 31.50 MiB | hist:
|
| 276 |
+
[ 211/ 291] blk.23.ffn_gate.weight - [ 4096, 14336, 1, 1], type = q8_0, quantizing to q3_K .. size = 59.50 MiB -> 24.06 MiB | hist:
|
| 277 |
+
[ 212/ 291] blk.23.ffn_up.weight - [ 4096, 14336, 1, 1], type = q8_0, quantizing to q3_K .. size = 59.50 MiB -> 24.06 MiB | hist:
|
| 278 |
+
[ 213/ 291] blk.23.ffn_norm.weight - [ 4096, 1, 1, 1], type = f32, size = 0.016 MB
|
| 279 |
+
[ 214/ 291] blk.23.attn_k.weight - [ 4096, 1024, 1, 1], type = q8_0, quantizing to q3_K .. size = 4.25 MiB -> 1.72 MiB | hist:
|
| 280 |
+
[ 215/ 291] blk.23.attn_output.weight - [ 4096, 4096, 1, 1], type = q8_0, quantizing to q4_K .. size = 17.00 MiB -> 9.00 MiB | hist:
|
| 281 |
+
[ 216/ 291] blk.23.attn_q.weight - [ 4096, 4096, 1, 1], type = q8_0, quantizing to q3_K .. size = 17.00 MiB -> 6.88 MiB | hist:
|
| 282 |
+
[ 217/ 291] blk.23.attn_v.weight - [ 4096, 1024, 1, 1], type = q8_0, quantizing to q4_K .. size = 4.25 MiB -> 2.25 MiB | hist:
|
| 283 |
+
[ 218/ 291] blk.24.attn_norm.weight - [ 4096, 1, 1, 1], type = f32, size = 0.016 MB
|
| 284 |
+
[ 219/ 291] blk.24.ffn_down.weight - [14336, 4096, 1, 1], type = q8_0, quantizing to q4_K .. size = 59.50 MiB -> 31.50 MiB | hist:
|
| 285 |
+
[ 220/ 291] blk.24.ffn_gate.weight - [ 4096, 14336, 1, 1], type = q8_0, quantizing to q3_K .. size = 59.50 MiB -> 24.06 MiB | hist:
|
| 286 |
+
[ 221/ 291] blk.24.ffn_up.weight - [ 4096, 14336, 1, 1], type = q8_0, quantizing to q3_K .. size = 59.50 MiB -> 24.06 MiB | hist:
|
| 287 |
+
[ 222/ 291] blk.24.ffn_norm.weight - [ 4096, 1, 1, 1], type = f32, size = 0.016 MB
|
| 288 |
+
[ 223/ 291] blk.24.attn_k.weight - [ 4096, 1024, 1, 1], type = q8_0, quantizing to q3_K .. size = 4.25 MiB -> 1.72 MiB | hist:
|
| 289 |
+
[ 224/ 291] blk.24.attn_output.weight - [ 4096, 4096, 1, 1], type = q8_0, quantizing to q4_K .. size = 17.00 MiB -> 9.00 MiB | hist:
|
| 290 |
+
[ 225/ 291] blk.24.attn_q.weight - [ 4096, 4096, 1, 1], type = q8_0, quantizing to q3_K .. size = 17.00 MiB -> 6.88 MiB | hist:
|
| 291 |
+
[ 226/ 291] blk.24.attn_v.weight - [ 4096, 1024, 1, 1], type = q8_0, quantizing to q4_K .. size = 4.25 MiB -> 2.25 MiB | hist:
|
| 292 |
+
[ 227/ 291] blk.25.attn_norm.weight - [ 4096, 1, 1, 1], type = f32, size = 0.016 MB
|
| 293 |
+
[ 228/ 291] blk.25.ffn_down.weight - [14336, 4096, 1, 1], type = q8_0, quantizing to q4_K .. size = 59.50 MiB -> 31.50 MiB | hist:
|
| 294 |
+
[ 229/ 291] blk.25.ffn_gate.weight - [ 4096, 14336, 1, 1], type = q8_0, quantizing to q3_K .. size = 59.50 MiB -> 24.06 MiB | hist:
|
| 295 |
+
[ 230/ 291] blk.25.ffn_up.weight - [ 4096, 14336, 1, 1], type = q8_0, quantizing to q3_K .. size = 59.50 MiB -> 24.06 MiB | hist:
|
| 296 |
+
[ 231/ 291] blk.25.ffn_norm.weight - [ 4096, 1, 1, 1], type = f32, size = 0.016 MB
|
| 297 |
+
[ 232/ 291] blk.25.attn_k.weight - [ 4096, 1024, 1, 1], type = q8_0, quantizing to q3_K .. size = 4.25 MiB -> 1.72 MiB | hist:
|
| 298 |
+
[ 233/ 291] blk.25.attn_output.weight - [ 4096, 4096, 1, 1], type = q8_0, quantizing to q4_K .. size = 17.00 MiB -> 9.00 MiB | hist:
|
| 299 |
+
[ 234/ 291] blk.25.attn_q.weight - [ 4096, 4096, 1, 1], type = q8_0, quantizing to q3_K .. size = 17.00 MiB -> 6.88 MiB | hist:
|
| 300 |
+
[ 235/ 291] blk.25.attn_v.weight - [ 4096, 1024, 1, 1], type = q8_0, quantizing to q4_K .. size = 4.25 MiB -> 2.25 MiB | hist:
|
| 301 |
+
[ 236/ 291] blk.26.attn_k.weight - [ 4096, 1024, 1, 1], type = q8_0, quantizing to q3_K .. size = 4.25 MiB -> 1.72 MiB | hist:
|
| 302 |
+
[ 237/ 291] blk.26.attn_output.weight - [ 4096, 4096, 1, 1], type = q8_0, quantizing to q4_K .. size = 17.00 MiB -> 9.00 MiB | hist:
|
| 303 |
+
[ 238/ 291] blk.26.attn_q.weight - [ 4096, 4096, 1, 1], type = q8_0, quantizing to q3_K .. size = 17.00 MiB -> 6.88 MiB | hist:
|
| 304 |
+
[ 239/ 291] blk.26.attn_v.weight - [ 4096, 1024, 1, 1], type = q8_0, quantizing to q4_K .. size = 4.25 MiB -> 2.25 MiB | hist:
|
| 305 |
+
[ 240/ 291] blk.26.attn_norm.weight - [ 4096, 1, 1, 1], type = f32, size = 0.016 MB
|
| 306 |
+
[ 241/ 291] blk.26.ffn_down.weight - [14336, 4096, 1, 1], type = q8_0, quantizing to q4_K .. size = 59.50 MiB -> 31.50 MiB | hist:
|
| 307 |
+
[ 242/ 291] blk.26.ffn_gate.weight - [ 4096, 14336, 1, 1], type = q8_0, quantizing to q3_K .. size = 59.50 MiB -> 24.06 MiB | hist:
|
| 308 |
+
[ 243/ 291] blk.26.ffn_up.weight - [ 4096, 14336, 1, 1], type = q8_0, quantizing to q3_K .. size = 59.50 MiB -> 24.06 MiB | hist:
|
| 309 |
+
[ 244/ 291] blk.26.ffn_norm.weight - [ 4096, 1, 1, 1], type = f32, size = 0.016 MB
|
| 310 |
+
[ 245/ 291] blk.27.attn_norm.weight - [ 4096, 1, 1, 1], type = f32, size = 0.016 MB
|
| 311 |
+
[ 246/ 291] blk.27.ffn_down.weight - [14336, 4096, 1, 1], type = q8_0, quantizing to q4_K .. size = 59.50 MiB -> 31.50 MiB | hist:
|
| 312 |
+
[ 247/ 291] blk.27.ffn_gate.weight - [ 4096, 14336, 1, 1], type = q8_0, quantizing to q3_K .. size = 59.50 MiB -> 24.06 MiB | hist:
|
| 313 |
+
[ 248/ 291] blk.27.ffn_up.weight - [ 4096, 14336, 1, 1], type = q8_0, quantizing to q3_K .. size = 59.50 MiB -> 24.06 MiB | hist:
|
| 314 |
+
[ 249/ 291] blk.27.ffn_norm.weight - [ 4096, 1, 1, 1], type = f32, size = 0.016 MB
|
| 315 |
+
[ 250/ 291] blk.27.attn_k.weight - [ 4096, 1024, 1, 1], type = q8_0, quantizing to q3_K .. size = 4.25 MiB -> 1.72 MiB | hist:
|
| 316 |
+
[ 251/ 291] blk.27.attn_output.weight - [ 4096, 4096, 1, 1], type = q8_0, quantizing to q4_K .. size = 17.00 MiB -> 9.00 MiB | hist:
|
| 317 |
+
[ 252/ 291] blk.27.attn_q.weight - [ 4096, 4096, 1, 1], type = q8_0, quantizing to q3_K .. size = 17.00 MiB -> 6.88 MiB | hist:
|
| 318 |
+
[ 253/ 291] blk.27.attn_v.weight - [ 4096, 1024, 1, 1], type = q8_0, quantizing to q4_K .. size = 4.25 MiB -> 2.25 MiB | hist:
|
| 319 |
+
[ 254/ 291] blk.28.attn_norm.weight - [ 4096, 1, 1, 1], type = f32, size = 0.016 MB
|
| 320 |
+
[ 255/ 291] blk.28.ffn_down.weight - [14336, 4096, 1, 1], type = q8_0, quantizing to q4_K .. size = 59.50 MiB -> 31.50 MiB | hist:
|
| 321 |
+
[ 256/ 291] blk.28.ffn_gate.weight - [ 4096, 14336, 1, 1], type = q8_0, quantizing to q3_K .. size = 59.50 MiB -> 24.06 MiB | hist:
|
| 322 |
+
[ 257/ 291] blk.28.ffn_up.weight - [ 4096, 14336, 1, 1], type = q8_0, quantizing to q3_K .. size = 59.50 MiB -> 24.06 MiB | hist:
|
| 323 |
+
[ 258/ 291] blk.28.ffn_norm.weight - [ 4096, 1, 1, 1], type = f32, size = 0.016 MB
|
| 324 |
+
[ 259/ 291] blk.28.attn_k.weight - [ 4096, 1024, 1, 1], type = q8_0, quantizing to q3_K .. size = 4.25 MiB -> 1.72 MiB | hist:
|
| 325 |
+
[ 260/ 291] blk.28.attn_output.weight - [ 4096, 4096, 1, 1], type = q8_0, quantizing to q4_K .. size = 17.00 MiB -> 9.00 MiB | hist:
|
| 326 |
+
[ 261/ 291] blk.28.attn_q.weight - [ 4096, 4096, 1, 1], type = q8_0, quantizing to q3_K .. size = 17.00 MiB -> 6.88 MiB | hist:
|
| 327 |
+
[ 262/ 291] blk.28.attn_v.weight - [ 4096, 1024, 1, 1], type = q8_0, quantizing to q4_K .. size = 4.25 MiB -> 2.25 MiB | hist:
|
| 328 |
+
[ 263/ 291] blk.29.attn_norm.weight - [ 4096, 1, 1, 1], type = f32, size = 0.016 MB
|
| 329 |
+
[ 264/ 291] blk.29.ffn_down.weight - [14336, 4096, 1, 1], type = q8_0, quantizing to q4_K .. size = 59.50 MiB -> 31.50 MiB | hist:
|
| 330 |
+
[ 265/ 291] blk.29.ffn_gate.weight - [ 4096, 14336, 1, 1], type = q8_0, quantizing to q3_K .. size = 59.50 MiB -> 24.06 MiB | hist:
|
| 331 |
+
[ 266/ 291] blk.29.ffn_up.weight - [ 4096, 14336, 1, 1], type = q8_0, quantizing to q3_K .. size = 59.50 MiB -> 24.06 MiB | hist:
|
| 332 |
+
[ 267/ 291] blk.29.ffn_norm.weight - [ 4096, 1, 1, 1], type = f32, size = 0.016 MB
|
| 333 |
+
[ 268/ 291] blk.29.attn_k.weight - [ 4096, 1024, 1, 1], type = q8_0, quantizing to q3_K .. size = 4.25 MiB -> 1.72 MiB | hist:
|
| 334 |
+
[ 269/ 291] blk.29.attn_output.weight - [ 4096, 4096, 1, 1], type = q8_0, quantizing to q4_K .. size = 17.00 MiB -> 9.00 MiB | hist:
|
| 335 |
+
[ 270/ 291] blk.29.attn_q.weight - [ 4096, 4096, 1, 1], type = q8_0, quantizing to q3_K .. size = 17.00 MiB -> 6.88 MiB | hist:
|
| 336 |
+
[ 271/ 291] blk.29.attn_v.weight - [ 4096, 1024, 1, 1], type = q8_0, quantizing to q4_K .. size = 4.25 MiB -> 2.25 MiB | hist:
|
| 337 |
+
[ 272/ 291] blk.30.ffn_gate.weight - [ 4096, 14336, 1, 1], type = q8_0, quantizing to q3_K .. size = 59.50 MiB -> 24.06 MiB | hist:
|
| 338 |
+
[ 273/ 291] blk.30.ffn_up.weight - [ 4096, 14336, 1, 1], type = q8_0, quantizing to q3_K .. size = 59.50 MiB -> 24.06 MiB | hist:
|
| 339 |
+
[ 274/ 291] blk.30.attn_k.weight - [ 4096, 1024, 1, 1], type = q8_0, quantizing to q3_K .. size = 4.25 MiB -> 1.72 MiB | hist:
|
| 340 |
+
[ 275/ 291] blk.30.attn_output.weight - [ 4096, 4096, 1, 1], type = q8_0, quantizing to q4_K .. size = 17.00 MiB -> 9.00 MiB | hist:
|
| 341 |
+
[ 276/ 291] blk.30.attn_q.weight - [ 4096, 4096, 1, 1], type = q8_0, quantizing to q3_K .. size = 17.00 MiB -> 6.88 MiB | hist:
|
| 342 |
+
[ 277/ 291] blk.30.attn_v.weight - [ 4096, 1024, 1, 1], type = q8_0, quantizing to q4_K .. size = 4.25 MiB -> 2.25 MiB | hist:
|
| 343 |
+
[ 278/ 291] output.weight - [ 4096, 32000, 1, 1], type = q8_0, quantizing to q6_K .. size = 132.81 MiB -> 102.54 MiB | hist:
|
| 344 |
+
[ 279/ 291] blk.30.attn_norm.weight - [ 4096, 1, 1, 1], type = f32, size = 0.016 MB
|
| 345 |
+
[ 280/ 291] blk.30.ffn_down.weight - [14336, 4096, 1, 1], type = q8_0, quantizing to q4_K .. size = 59.50 MiB -> 31.50 MiB | hist:
|
| 346 |
+
[ 281/ 291] blk.30.ffn_norm.weight - [ 4096, 1, 1, 1], type = f32, size = 0.016 MB
|
| 347 |
+
[ 282/ 291] blk.31.attn_norm.weight - [ 4096, 1, 1, 1], type = f32, size = 0.016 MB
|
| 348 |
+
[ 283/ 291] blk.31.ffn_down.weight - [14336, 4096, 1, 1], type = q8_0, quantizing to q4_K .. size = 59.50 MiB -> 31.50 MiB | hist:
|
| 349 |
+
[ 284/ 291] blk.31.ffn_gate.weight - [ 4096, 14336, 1, 1], type = q8_0, quantizing to q3_K .. size = 59.50 MiB -> 24.06 MiB | hist:
|
| 350 |
+
[ 285/ 291] blk.31.ffn_up.weight - [ 4096, 14336, 1, 1], type = q8_0, quantizing to q3_K .. size = 59.50 MiB -> 24.06 MiB | hist:
|
| 351 |
+
[ 286/ 291] blk.31.ffn_norm.weight - [ 4096, 1, 1, 1], type = f32, size = 0.016 MB
|
| 352 |
+
[ 287/ 291] blk.31.attn_k.weight - [ 4096, 1024, 1, 1], type = q8_0, quantizing to q3_K .. size = 4.25 MiB -> 1.72 MiB | hist:
|
| 353 |
+
[ 288/ 291] blk.31.attn_output.weight - [ 4096, 4096, 1, 1], type = q8_0, quantizing to q4_K .. size = 17.00 MiB -> 9.00 MiB | hist:
|
| 354 |
+
[ 289/ 291] blk.31.attn_q.weight - [ 4096, 4096, 1, 1], type = q8_0, quantizing to q3_K .. size = 17.00 MiB -> 6.88 MiB | hist:
|
| 355 |
+
[ 290/ 291] blk.31.attn_v.weight - [ 4096, 1024, 1, 1], type = q8_0, quantizing to q4_K .. size = 4.25 MiB -> 2.25 MiB | hist:
|
| 356 |
+
[ 291/ 291] output_norm.weight - [ 4096, 1, 1, 1], type = f32, size = 0.016 MB
|
| 357 |
+
llama_model_quantize_internal: model size = 7338.64 MB
|
| 358 |
+
llama_model_quantize_internal: quant size = 3355.27 MB
|
| 359 |
+
|
| 360 |
+
main: quantize time = 165764.53 ms
|
| 361 |
+
main: total time = 165764.53 ms
|
| 362 |
+
|