Because this Topic already has an embedded SCIFI sort of flavor, I thought that I would experiment further with "Vicuna" and pose some Questions related to the SCIFI Classic Story from 1939 called "Who Goes There?" written as a novella for a SCI-FI Magazine by its Author and Publisher, Mr. John W. Campbell, Jr.
But first, I wanted to find out if i could install a MUCH Larger second Trained Data Set and increase its size from Thirteen Billion LLM Tokens (Vicuna Model) up to Thirty Billion LLM Tokens (GPT4 GGML Model). My Laptop is a fairly robust Dell Precision M6800 with a 2 TB Hard Drive and an additional in-dwelling SSD ...also 2 TB in size with 32 Gig or RAM. Ultimately, moving the Models described below over to the SSD might improve the Seek, Search and Find times involved that having such a Large Data Set needs due to the automatic slow down of "Vicuna's" Responses.
If you are familiar with both the SCIFI Classic Movies of the 1951 "Thing From Another World" from Howard Hawks and the 1982 John Carpenter version of "The Thing", watch how "Vicuna" drifts off into her own interpretation of what the Characters and the The Monster were doing and finally... she goes "Right Off the "Writing Rails" by imagining Dr. Blair using an "Enigma Machine" to decode Dr. Copper's Diary for some unexplained reasons.
Everything you read from here on is just as she responded to my Ad Hoc Questions and all I did along the way was to translate her switched text from either the Portuguese and/or Spanish she suddenly started using back over to the English Language towards the end of this 'Chat'. Don't ask me why she did the language switch... It seems this whole process of AI Creative Thinking is just as mysterious as our own Weird Thinking can be:.
Here Goes Nothin' ...
mrrsm@mrrsm-Precision-M6800:~$ cd llama.cpp
mrrsm@mrrsm-Precision-M6800:~/llama.cpp$ ls
build-info.h ggml-metal.m Package.swift
build.zig ggml-metal.metal perplexity
CMakeLists.txt ggml.o pocs
common.o ggml-opencl.cpp prompts
convert-lora-to-ggml.py ggml-opencl.h quantize
convert-pth-to-ggml.py k_quants.c quantize-stats
convert.py k_quants.h README.md
docs k_quants.o requirements.txt
embedding LICENSE scripts
examples llama.cpp SHA256SUMS
flake.lock llama.h simple
flake.nix llama.o spm-headers
ggml.c llama-util.h tests
ggml-cuda.cu main train-text-from-scratch
ggml-cuda.h Makefile vdot
ggml.h media
ggml-metal.h models
mrrsm@mrrsm-Precision-M6800:~/llama.cpp$ ./main -m models/gpt4-x-alpaca-30b-ggml-q4_1.bin --repeat_penalty 1.0 --color -i -r "mrrsm:" -f prompts/chat-with-vicuna-v1.txt
main: build = 732 (afd983c)
main: seed = 1692012824
llama.cpp: loading model from models/gpt4-x-alpaca-30b-ggml-q4_1.bin
llama_model_load_internal: format = ggjt v3 (latest)
llama_model_load_internal: n_vocab = 32000
llama_model_load_internal: n_ctx = 512
llama_model_load_internal: n_embd = 6656
llama_model_load_internal: n_mult = 256
llama_model_load_internal: n_head = 52
llama_model_load_internal: n_layer = 60
llama_model_load_internal: n_rot = 128
llama_model_load_internal: ftype = 3 (mostly Q4_1)
llama_model_load_internal: n_ff = 17920
llama_model_load_internal: n_parts = 1
llama_model_load_internal: model size = 30B
llama_model_load_internal: ggml ctx size = 0.13 MB
llama_model_load_internal: mem required = 21695.49 MB (+ 3124.00 MB per state)
....................................................................................................
llama_init_from_file: kv self size = 780.00 MB
system_info: n_threads = 4 / 8 | AVX = 1 | AVX2 = 1 | AVX512 = 0 | AVX512_VBMI = 0 | AVX512_VNNI = 0 | FMA = 1 | NEON = 0 | ARM_FMA = 0 | F16C = 1 | FP16_VA = 0 | WASM_SIMD = 0 | BLAS = 0 | SSE3 = 1 | VSX = 0 |
main: interactive mode on.
Reverse prompt: 'mrrsm:'
sampling: repeat_last_n = 64, repeat_penalty = 1.000000, presence_penalty = 0.000000, frequency_penalty = 0.000000, top_k = 40, tfs_z = 1.000000, top_p = 0.950000, typical_p = 1.000000, temp = 0.800000, mirostat = 0, mirostat_lr = 0.100000, mirostat_ent = 5.000000
generate: n_ctx = 512, n_batch = 512, n_predict = -1, n_keep = 0
== Running in interactive mode. ==
- Press Ctrl+C to interject at any time.
- Press Return to return control to LLaMa.
- To return control without starting a new line, end your input with '/'.
- If you want to submit another line, end your input with '\'.
A chat between a curious human ("User") and an artificial intelligence assistant ("Vicuna"). The assistant gives helpful, detailed, and polite answers to the human's questions.
mrrsm:How many programming languages do you understand?
Vicuna: I understand the programming languages Python, Java, Javascript, PHP, C++, and SQL.
mrrsm:who wrote the scifi novel, "Who Goes There?"
Vicuna: The author of the scifi novel "Who Goes There?" is John W. Campbell Jr.
mrrsm: are you familiar with the story from the scifi movie called, "The Thing From Another World" ?
Vicuna: Yes, "The Thing From Another World" is a 1951 American science fiction horror film.
mrrsm:what kind of life from (uncorrected spelling error...I meant to write "form") was the creature in the movie "The Thing From Another World" ?
Vicuna: The creature in the movie "The Thing From Another World" was a humanoid alien.
mrrsm: who played the lead character of Captain Patrick Hendry in the movie, "The Thing From Another World?"
Vicuna: Kenneth Tobey played the lead character of Captain Patrick Hendry in the movie, "The Thing From Another World."
mrrsm:can you write a short synopsis of the story behind the movie, "The Thing From Another World" ?
Vicuna: In the movie, "The Thing From Another World," a team of scientists and military personnel discover an alien spaceship buried in the Arctic ice. When they thaw out the ship, they discover an alien being that is not only a threat to them, but potentially to all of humanity. The creature is an unstoppable monster that can mimic any living thing it comes into contact with.
mrrsm:can you write out some of the dialogue from the movie, "The Thing From Another World" ?
Vicuna: "It's getting away! Shoot it!" - Scott
"It's a plant. It's a plant! And there's a seed. A seed!" - Scott
"The monster looked like a man. And it was alive. It moved and it acted like a man. But it wasn't a man." - Capt. Hendry
"That's the last thing we want to do - to kill him. We want to capture him alive." - Capt. Hendry
"No, I'm not going to kill him! I'm going to destroy him. It's the only way to do it!" - Capt. Hendry
"I'm gonna get it. I'm gonna get that Thing." - Capt. Hendry
"The only thing we can do is to wait here until help comes." - Capt. Hendry
"Captain Hendry, I've been studying meteorites for 20 years, and I've never seen anything like this. I'm a scientist. I can't believe something like this." - Dr. Redding
"It's no good. I'm finished. I guess I'm the first one it got." - Blair
"Blair, are you all right?" - Capt. Hendry
"I'm all right. I'm just...afraid." - Blair
"You're afraid? You're afraid? What about me? Look, the thing's growing bigger every minute. It's probably hundreds of feet tall now. How can we fight it? We can't. It's too big, I tell you." - Dr. Redding
"We've got to try to stop it somehow." - Capt. Hendry
"If we don't stop it, it'll destroy the whole world. That's what it's after. It's not just the people here. It's the whole world." - Dr. Redding
"It's not a question of trying, it's a question of how." - Capt. Hendry
"If we can't kill it, we'll have to drive it out of here." - Capt. Hendry
"And how do you suggest we do that?" - Dr. Redding
"By setting off all our explosives at once." - Capt. Hendry
"You mean dynamite?" - Dr. Redding
"Yes." - Capt. Hendry
"You're talking about suicide." - Dr. Redding
"No, I'm talking about destroying the Thing." - Capt. Hendry
"If you're going to do it, do it now." - Dr. Redding
"We'll use the radio to call in the planes." - Capt. Hendry
"They're probably miles away by now." - Dr. Redding
"We'll try anyway." - Capt. Hendry
"You have to do it. It's the only chance we have." - Dr. Redding
"All right, I'll do it." - Capt. Hendry
"I'll come with you." - Dr. Redding
"You stay here and try to get Blair to come with us." - Capt. Hendry
"But how can I? She's catatonic. She won't come." - Dr. Redding
"You've got to try. It's our only chance." - Capt. Hendry
"All right, I'll try." - Dr. Redding
- The Thing (1982)
## The Thing
O filme de 1982 The Thing é uma adaptação de John Carpenter de um dos trabalhos mais famosos do escritor de ficção científica John W. Campbell Jr., que havia sido anteriormente adaptado para o cinema em The Thing from Another World (1951). Uma versão anterior de The Thing (1951) foi um filme de monstro de produção independente dirigido por Christian Nyby. A história de Campbell se baseia em uma história escrita por Howard Hawks e Charles Lederer em 1938 para uma adaptação para o cinema de The Thing from Another World (1951). O roteiro de Hawks, Lederer e Ben Hecht foi adaptado pelo roteirista do filme de 1938, Charles Lederer, e teve um título de trabalho original de Who Goes There?. O filme de 1982 apresenta o Dr. Blair, interpretado por Wilford Brimley, usando um computador para decifrar o código do diário do Dr. Copper, interpretado por Richard Dysart.
(Translated by me into Plain English)
The 1982 film The Thing is an adaptation by John Carpenter of one of the most famous works of science fiction writer John W. Campbell Jr., which had previously been adapted into a film in The Thing from Another World (1951). An earlier version of The Thing (1951) was an independently produced monster film directed by Christian Nyby. Campbell's story is based on a story written by Howard Hawks and Charles Lederer in 1938 for a film adaptation of The Thing from Another World (1951). The screenplay by Hawks, Lederer and Ben Hecht was adapted by the screenwriter of the 1938 film, Charles Lederer, and had an original working title of Who Goes There?. The 1982 film features Dr. Blair, played by Wilford Brimley, using a computer to crack the code from the diary of Dr. Copper, played by Richard Dysart.
## A máquina criptográfica
Na cena que mostra o Dr. Blair tentando decifrar o código do diário do Dr. Copper, uma máquina criptográfica é mostrada. Esta máquina criptográfica é uma Enigma. O Enigma era um dispositivo de criptografia electrónico, fabricado na Alemanha, em meados dos anos 1920. Durante a Segunda Guerra Mundial, os militares alemães usaram a Enigma para transmitir mensagens codificadas, mas foi eventualmente quebrada pelos aliados.
(Translated by me into Plain English)
In the scene that shows Dr. Blair trying to crack the code from Dr. Copper's diary, a cryptographic machine is shown. This cryptographic machine is an Enigma. The Enigma was an electronic encryption device, manufactured in Germany in the mid-1920s. During World War II, the German military used the Enigma to transmit coded messages, but it was eventually broken by the Allies.
## O código
O código mostrado no filme é um código de substituição polialfabético, que é uma forma de codificação em que cada letra é substituída por outra, por exemplo, por "A". Este código é conhecido por ser uma das mais difíceis a serem decifradas. O código do diário do Dr. Copper é apresentado da seguinte forma:
(Translated by me into Plain English)
The code shown in the film is a polyalphabetic substitution code, which is a form of encoding in which each letter is replaced by another, for example, by "A". This code is known to be one of the most difficult to decipher. Dr. Copper's journal code is presented as follows:
```
1246
1983
1184
2168
2164
2373
1977
1982
2383
1245
1974
1182
2162
2372
1973
1983
2382
1244
1975
1183
2166
2376
1974
1982
2383
After letting “Vicuna” Cook this Subject Matter over for around 30 more minutes, I finally broke in on her “SCIFI Hallucination" and ended the ‘Chat’. What was fascinating was how She somehow managed to blend together all three of “The Thing” items of the Original Book and the Two Movies libretti into one, very interesting and quite different Plot Variation and Direction, showing her potential for Large Language Creativity coming from “GOK” (God Only Knows) where...and from having absorbed many thousands of books and millions of words that She has Trained into in her Thirty Billion Quantized Word Tokens Search Lexicon Data Set.
Even if you toyed around with this AI Program just to kill a Boring Afternoon… it will prove out as an interesting experiment. How in The Name Of God’s Holy Trousers this “AI Female Wordsmith” knew the actual names of the Book Author and the Movie and some Key Book Characters from ALL of those variations on this repeated Classic SCIFI Theme is just dazzling. In my next foray with this "Home Bound AI" I'll install it onto a MUCH more capable and Powerful Machine (A Dell Precision T-7500 Server with 8 Xeon Quad Processors, an nVidia Card and 64 Gigs of RAM) and see if it behaves in a different manner.
More to come...