|
22 | 22 | "source": [
|
23 | 23 | "## Download a GGUF Model\n",
|
24 | 24 | "\n",
|
25 |
| - "Lets download a GGUF model to test it out. For this, we will use [microsoft/Phi-3-mini-4k-instruct-gguf](https://huggingface.co/microsoft/Phi-3-mini-4k-instruct-gguf). It is a 3.8B parameter model which also is available in 4-bit quantization. \n", |
| 25 | + "Lets download a GGUF model to test it out. For this, we will use [bartowski/Phi-3.5-mini-instruct-GGUF](https://huggingface.co/bartowski/Phi-3.5-mini-instruct-GGUF). It is a 3.8B parameter model which also is available in 4-bit quantization. \n", |
26 | 26 | "\n",
|
27 | 27 | "We can download the model by selecting the q4 GGUF file from the \"Files and versions\" tab.\n",
|
28 | 28 | "\n",
|
|
38 | 38 | "name": "stdout",
|
39 | 39 | "output_type": "stream",
|
40 | 40 | "text": [
|
41 |
| - "--2024-07-20 11:11:30-- https://huggingface.co/microsoft/Phi-3-mini-4k-instruct-gguf/resolve/main/Phi-3-mini-4k-instruct-q4.gguf?download=true\n", |
| 41 | + "--2024-07-20 11:11:30-- https://huggingface.co/bartowski/Phi-3.5-mini-instruct-GGUF/resolve/main/Phi-3.5-mini-instruct-Q4_K_M.gguf?download=true\n", |
42 | 42 | "Resolving huggingface.co (huggingface.co)... 2600:9000:275f:7600:17:b174:6d00:93a1, 2600:9000:275f:3800:17:b174:6d00:93a1, 2600:9000:275f:6e00:17:b174:6d00:93a1, ...\n",
|
43 | 43 | "Connecting to huggingface.co (huggingface.co)|2600:9000:275f:7600:17:b174:6d00:93a1|:443... connected.\n",
|
44 | 44 | "HTTP request sent, awaiting response... 302 Found\n",
|
45 |
| - "Location: https://cdn-lfs-us-1.huggingface.co/repos/41/c8/41c860f65b01de5dc4c68b00d84cead799d3e7c48e38ee749f4c6057776e2e9e/8a83c7fb9049a9b2e92266fa7ad04933bb53aa1e85136b7b30f1b8000ff2edef?response-content-disposition=attachment%3B+filename*%3DUTF-8%27%27Phi-3-mini-4k-instruct-q4.gguf%3B+filename%3D%22Phi-3-mini-4k-instruct-q4.gguf%22%3B&Expires=1721725890&Policy=eyJTdGF0ZW1lbnQiOlt7IkNvbmRpdGlvbiI6eyJEYXRlTGVzc1RoYW4iOnsiQVdTOkVwb2NoVGltZSI6MTcyMTcyNTg5MH19LCJSZXNvdXJjZSI6Imh0dHBzOi8vY2RuLWxmcy11cy0xLmh1Z2dpbmdmYWNlLmNvL3JlcG9zLzQxL2M4LzQxYzg2MGY2NWIwMWRlNWRjNGM2OGIwMGQ4NGNlYWQ3OTlkM2U3YzQ4ZTM4ZWU3NDlmNGM2MDU3Nzc2ZTJlOWUvOGE4M2M3ZmI5MDQ5YTliMmU5MjI2NmZhN2FkMDQ5MzNiYjUzYWExZTg1MTM2YjdiMzBmMWI4MDAwZmYyZWRlZj9yZXNwb25zZS1jb250ZW50LWRpc3Bvc2l0aW9uPSoifV19&Signature=joXQf4QRpEhtFeQ3r3gJ0zyJ3bXReb9OxM%7EZit3GJ3355ycKQzemJ%7E6eD-J7%7EkphnsPpRpUDhQkCr2-Oidqo7dgltmFsWAX4SmQLn65R1yjO%7EsMvi%7E4vOUpaRPYlSMCyWWJpiZZjQYVH4Uk0o-G62ALFXKGaDfr627kvahP-fJYwNNP1riTrH8hbbah28ZKRAQjUGI1aNqerG0jojudnGOagawISAnudkAOFZfxnN7Qw3CoMySZLj9Euu02RBv2A5Yy0uSjG7b8rilx-tU5HDR3ECohdQQ8yPXjYFU-LZi-zcG1wwBDF-S01qb%7EgPWsTorenxfRM2cG6J%7EvSziGCzA__&Key-Pair-Id=K24J24Z295AEI9 [following]\n", |
46 |
| - "--2024-07-20 11:11:30-- https://cdn-lfs-us-1.huggingface.co/repos/41/c8/41c860f65b01de5dc4c68b00d84cead799d3e7c48e38ee749f4c6057776e2e9e/8a83c7fb9049a9b2e92266fa7ad04933bb53aa1e85136b7b30f1b8000ff2edef?response-content-disposition=attachment%3B+filename*%3DUTF-8%27%27Phi-3-mini-4k-instruct-q4.gguf%3B+filename%3D%22Phi-3-mini-4k-instruct-q4.gguf%22%3B&Expires=1721725890&Policy=eyJTdGF0ZW1lbnQiOlt7IkNvbmRpdGlvbiI6eyJEYXRlTGVzc1RoYW4iOnsiQVdTOkVwb2NoVGltZSI6MTcyMTcyNTg5MH19LCJSZXNvdXJjZSI6Imh0dHBzOi8vY2RuLWxmcy11cy0xLmh1Z2dpbmdmYWNlLmNvL3JlcG9zLzQxL2M4LzQxYzg2MGY2NWIwMWRlNWRjNGM2OGIwMGQ4NGNlYWQ3OTlkM2U3YzQ4ZTM4ZWU3NDlmNGM2MDU3Nzc2ZTJlOWUvOGE4M2M3ZmI5MDQ5YTliMmU5MjI2NmZhN2FkMDQ5MzNiYjUzYWExZTg1MTM2YjdiMzBmMWI4MDAwZmYyZWRlZj9yZXNwb25zZS1jb250ZW50LWRpc3Bvc2l0aW9uPSoifV19&Signature=joXQf4QRpEhtFeQ3r3gJ0zyJ3bXReb9OxM%7EZit3GJ3355ycKQzemJ%7E6eD-J7%7EkphnsPpRpUDhQkCr2-Oidqo7dgltmFsWAX4SmQLn65R1yjO%7EsMvi%7E4vOUpaRPYlSMCyWWJpiZZjQYVH4Uk0o-G62ALFXKGaDfr627kvahP-fJYwNNP1riTrH8hbbah28ZKRAQjUGI1aNqerG0jojudnGOagawISAnudkAOFZfxnN7Qw3CoMySZLj9Euu02RBv2A5Yy0uSjG7b8rilx-tU5HDR3ECohdQQ8yPXjYFU-LZi-zcG1wwBDF-S01qb%7EgPWsTorenxfRM2cG6J%7EvSziGCzA__&Key-Pair-Id=K24J24Z295AEI9\n", |
| 45 | + "Location: https://cdn-lfs-us-1.huggingface.co/repos/41/c8/41c860f65b01de5dc4c68b00d84cead799d3e7c48e38ee749f4c6057776e2e9e/8a83c7fb9049a9b2e92266fa7ad04933bb53aa1e85136b7b30f1b8000ff2edef?response-content-disposition=attachment%3B+filename*%3DUTF-8%27%27Phi-3.5-mini-instruct-Q4_K_M.gguf%3B+filename%3D%22P Phi-3.5-mini-instruct-Q4_K_M.gguf%22%3B&Expires=1721725890&Policy=eyJTdGF0ZW1lbnQiOlt7IkNvbmRpdGlvbiI6eyJEYXRlTGVzc1RoYW4iOnsiQVdTOkVwb2NoVGltZSI6MTcyMTcyNTg5MH19LCJSZXNvdXJjZSI6Imh0dHBzOi8vY2RuLWxmcy11cy0xLmh1Z2dpbmdmYWNlLmNvL3JlcG9zLzQxL2M4LzQxYzg2MGY2NWIwMWRlNWRjNGM2OGIwMGQ4NGNlYWQ3OTlkM2U3YzQ4ZTM4ZWU3NDlmNGM2MDU3Nzc2ZTJlOWUvOGE4M2M3ZmI5MDQ5YTliMmU5MjI2NmZhN2FkMDQ5MzNiYjUzYWExZTg1MTM2YjdiMzBmMWI4MDAwZmYyZWRlZj9yZXNwb25zZS1jb250ZW50LWRpc3Bvc2l0aW9uPSoifV19&Signature=joXQf4QRpEhtFeQ3r3gJ0zyJ3bXReb9OxM%7EZit3GJ3355ycKQzemJ%7E6eD-J7%7EkphnsPpRpUDhQkCr2-Oidqo7dgltmFsWAX4SmQLn65R1yjO%7EsMvi%7E4vOUpaRPYlSMCyWWJpiZZjQYVH4Uk0o-G62ALFXKGaDfr627kvahP-fJYwNNP1riTrH8hbbah28ZKRAQjUGI1aNqerG0jojudnGOagawISAnudkAOFZfxnN7Qw3CoMySZLj9Euu02RBv2A5Yy0uSjG7b8rilx-tU5HDR3ECohdQQ8yPXjYFU-LZi-zcG1wwBDF-S01qb%7EgPWsTorenxfRM2cG6J%7EvSziGCzA__&Key-Pair-Id=K24J24Z295AEI9 [following]\n", |
| 46 | + "--2024-07-20 11:11:30-- https://cdn-lfs-us-1.huggingface.co/repos/41/c8/41c860f65b01de5dc4c68b00d84cead799d3e7c48e38ee749f4c6057776e2e9e/8a83c7fb9049a9b2e92266fa7ad04933bb53aa1e85136b7b30f1b8000ff2edef?response-content-disposition=attachment%3B+filename*%3DUTF-8%27%27Phi-3.5-mini-instruct-Q4_K_M.gguf%3B+filename%3D%22P Phi-3.5-mini-instruct-Q4_K_M.gguf%22%3B&Expires=1721725890&Policy=eyJTdGF0ZW1lbnQiOlt7IkNvbmRpdGlvbiI6eyJEYXRlTGVzc1RoYW4iOnsiQVdTOkVwb2NoVGltZSI6MTcyMTcyNTg5MH19LCJSZXNvdXJjZSI6Imh0dHBzOi8vY2RuLWxmcy11cy0xLmh1Z2dpbmdmYWNlLmNvL3JlcG9zLzQxL2M4LzQxYzg2MGY2NWIwMWRlNWRjNGM2OGIwMGQ4NGNlYWQ3OTlkM2U3YzQ4ZTM4ZWU3NDlmNGM2MDU3Nzc2ZTJlOWUvOGE4M2M3ZmI5MDQ5YTliMmU5MjI2NmZhN2FkMDQ5MzNiYjUzYWExZTg1MTM2YjdiMzBmMWI4MDAwZmYyZWRlZj9yZXNwb25zZS1jb250ZW50LWRpc3Bvc2l0aW9uPSoifV19&Signature=joXQf4QRpEhtFeQ3r3gJ0zyJ3bXReb9OxM%7EZit3GJ3355ycKQzemJ%7E6eD-J7%7EkphnsPpRpUDhQkCr2-Oidqo7dgltmFsWAX4SmQLn65R1yjO%7EsMvi%7E4vOUpaRPYlSMCyWWJpiZZjQYVH4Uk0o-G62ALFXKGaDfr627kvahP-fJYwNNP1riTrH8hbbah28ZKRAQjUGI1aNqerG0jojudnGOagawISAnudkAOFZfxnN7Qw3CoMySZLj9Euu02RBv2A5Yy0uSjG7b8rilx-tU5HDR3ECohdQQ8yPXjYFU-LZi-zcG1wwBDF-S01qb%7EgPWsTorenxfRM2cG6J%7EvSziGCzA__&Key-Pair-Id=K24J24Z295AEI9\n", |
47 | 47 | "Resolving cdn-lfs-us-1.huggingface.co (cdn-lfs-us-1.huggingface.co)... 2600:9000:275f:7a00:17:9a40:4dc0:93a1, 2600:9000:275f:fc00:17:9a40:4dc0:93a1, 2600:9000:275f:4800:17:9a40:4dc0:93a1, ...\n",
|
48 | 48 | "Connecting to cdn-lfs-us-1.huggingface.co (cdn-lfs-us-1.huggingface.co)|2600:9000:275f:7a00:17:9a40:4dc0:93a1|:443... connected.\n",
|
49 | 49 | "HTTP request sent, awaiting response... 200 OK\n",
|
50 | 50 | "Length: 2393231072 (2.2G) [binary/octet-stream]\n",
|
51 |
| - "Saving to: ‘Phi-3-mini-4k-instruct-q4.gguf?download=true’\n", |
| 51 | + "Saving to: ‘Phi-3.5-mini-instruct-Q4_K_M.gguf?download=true’\n", |
52 | 52 | "\n",
|
53 | 53 | "Phi-3-mini-4k-instr 100%[===================>] 2.23G 22.5MB/s in 96s \n",
|
54 | 54 | "\n",
|
55 |
| - "2024-07-20 11:13:06 (23.7 MB/s) - ‘Phi-3-mini-4k-instruct-q4.gguf?download=true’ saved [2393231072/2393231072]\n", |
| 55 | + "2024-07-20 11:13:06 (23.7 MB/s) - ‘Phi-3.5-mini-instruct-Q4_K_M.gguf?download=true’ saved [2393231072/2393231072]\n", |
56 | 56 | "\n"
|
57 | 57 | ]
|
58 | 58 | }
|
59 | 59 | ],
|
60 | 60 | "source": [
|
61 |
| - "EXPORT_PATH = \"Phi-3-mini-4k-instruct-q4.gguf\"\n", |
62 |
| - "! wget \"https://huggingface.co/microsoft/Phi-3-mini-4k-instruct-gguf/resolve/main/Phi-3-mini-4k-instruct-q4.gguf?download=true\" -O {EXPORT_PATH}" |
| 61 | + "EXPORT_PATH = \"Phi-3.5-mini-instruct-Q4_K_M.gguf\"\n", |
| 62 | + "! wget \"https://huggingface.co/bartowski/Phi-3.5-mini-instruct-GGUF/resolve/main/Phi-3.5-mini-instruct-Q4_K_M.gguf?download=true\" -O {EXPORT_PATH}" |
63 | 63 | ]
|
64 | 64 | },
|
65 | 65 | {
|
|
359 | 359 | "metadata": {},
|
360 | 360 | "outputs": [],
|
361 | 361 | "source": [
|
362 |
| - "autoGGUFModel.write().overwrite().save(f\"Phi-3-mini-4k-instruct-q4_spark_nlp\")" |
| 362 | + "autoGGUFModel.write().overwrite().save(f\"Phi-3.5-mini-instruct-Q4_K_M_spark_nlp\")" |
363 | 363 | ]
|
364 | 364 | },
|
365 | 365 | {
|
|
398 | 398 | "text": [
|
399 | 399 | "total 2337168\n",
|
400 | 400 | "drwxr-xr-x 2 ducha ducha 4096 Jul 21 16:24 metadata\n",
|
401 |
| - "-rwxrwxr-x 1 ducha ducha 2393231072 Jul 21 16:24 Phi-3-mini-4k-instruct-q4.gguf\n" |
| 401 | + "-rwxrwxr-x 1 ducha ducha 2393231072 Jul 21 16:24 Phi-3.5-mini-instruct-Q4_K_M.gguf\n" |
402 | 402 | ]
|
403 | 403 | }
|
404 | 404 | ],
|
405 | 405 | "source": [
|
406 |
| - "! ls -l Phi-3-mini-4k-instruct-q4_spark_nlp" |
| 406 | + "! ls -l Phi-3.5-mini-instruct-Q4_K_M_spark_nlp" |
407 | 407 | ]
|
408 | 408 | },
|
409 | 409 | {
|
|
437 | 437 | "name": "stderr",
|
438 | 438 | "output_type": "stream",
|
439 | 439 | "text": [
|
440 |
| - "llama_model_loader: loaded meta data with 24 key-value pairs and 195 tensors from /tmp/spark-bbad4f64-91a7-4b6e-8242-7f91e6abca54/userFiles-f7d4e4e9-c02d-46e4-81b5-bf5a26d70930/Phi-3-mini-4k-instruct-q4.gguf (version GGUF V3 (latest))\n", |
| 440 | + "llama_model_loader: loaded meta data with 24 key-value pairs and 195 tensors from /tmp/spark-bbad4f64-91a7-4b6e-8242-7f91e6abca54/userFiles-f7d4e4e9-c02d-46e4-81b5-bf5a26d70930/Phi-3.5-mini-instruct-Q4_K_M.gguf (version GGUF V3 (latest))\n", |
441 | 441 | "llama_model_loader: Dumping metadata keys/values. Note: KV overrides do not apply in this output.\n",
|
442 | 442 | "llama_model_loader: - kv 0: general.architecture str = phi3\n",
|
443 | 443 | "llama_model_loader: - kv 1: general.name str = Phi3\n",
|
|
583 | 583 | "\n",
|
584 | 584 | "document_assembler = DocumentAssembler().setInputCol(\"text\").setOutputCol(\"document\")\n",
|
585 | 585 | "\n",
|
586 |
| - "auto_gguf_model = AutoGGUFModel.load(\"Phi-3-mini-4k-instruct-q4_spark_nlp\")\n", |
| 586 | + "auto_gguf_model = AutoGGUFModel.load(\"Phi-3.5-mini-instruct-Q4_K_M_spark_nlp\")\n", |
587 | 587 | "\n",
|
588 | 588 | "pipeline = Pipeline().setStages([document_assembler, auto_gguf_model])\n",
|
589 | 589 | "\n",
|
|
0 commit comments